Ultimate Guide To Data Science Interviews

User Manual:

Open the PDF directly: View PDF PDF.
Page Count: 87

Table of Contents
TableofContents
Introduction
WhatisDataScience?
DifferentRoleswithinDataScience
HowDifferentCompaniesThinkAboutDataScience
1Earlystagestartups(200employeesorfewer)lookingtobuildadataproduct
2Earlystagestartups(200employeesorfewer)lookingtotakeadvantageoftheirdata
3MidsizeandlargeFortune500companieswhoarelookingtotakeadvantageoftheir
data
4Largetechnologycompanieswithwellestablisheddatateams
IndustriesthatemployDataScientists
GettingaDataScienceInterview
NinePathstoaDataScienceInterview
TraditionalPathstoJobInterviews
1DataScienceJobBoardsandStandardJobApplications
2WorkwithaRecruiter
3GotoJobFairs
ProactivePathstoJobInterviews
4AttendorOrganizeaDataScienceEvent
5FreelanceandBuildaPortfolio
6GetInvolvedinOpenDataandOpenSource
7ParticipateinDataScienceCompetitions
8AskforCoffees,doInformationalInterviews
9AttendDataHackathons
WorkingwithRecruiters
HowtoApply
CVvsLinkedIn
CoverLettervsEmail
www.springboard.com
ULTIMATE GUIDE TO DATA SCIENCE INTERVIEWS
1
HowtogetReferencesandYourNetworktoWorkforYou
PreparingfortheInterview
WhattoExpect
1ThePhoneScreen
2TakehomeAssignment
3PhoneCallwithaHiringManager
4OnsiteInterviewwithaHiringManager
5TechnicalChallenge
6InterviewwithanExecutive
Whatadatascientistisbeingevaluatedon
TheCategoriesofDataScienceQuestions
BehavioralQuestions
MathematicsQuestions
StatisticsQuestions
ScenarioQuestions
TacklingtheInterview
Conclusion
WhatHiringManagersareLookingFor
InterviewwithWillKurt(QuickSprout)
InterviewwithMattFornito(OpsVisionSolutions)
InterviewwithAndrewMaguire(PMC/Google/Accenture)
InterviewwithHristoGyoshev(MasterClass)
Conclusion
HowSuccessfulIntervieweesMadeIt
SaraWeinstein
NirajSheth
SdrjanSantic
Conclusion
7ThingstoDoAfterTheInterview
1Sendafollowupthankyounote
2Sendthemthoughtsonsomethingtheybroughtupintheinterview
3Sendrelevantwork/homeworktotheemployer
www.springboard.com
ULTIMATE GUIDE TO DATA SCIENCE INTERVIEWS
2
4Keepintouch,therightway
5Leverageconnections
6Acceptanyrejectionwithprofessionalism
7Keepuphope
TheOfferProcess
HandlingOffers
CompanyCulture
Team
Location
NegotiatingYourSalary
FactsandFigures
TakingtheOffertotheBestFirstDay
Templates
Reachingouttogetareferral
Followingupafteraninterview
Resources
 
www.springboard.com
ULTIMATE GUIDE TO DATA SCIENCE INTERVIEWS
3
Introduction
WhenwefirstwrotetheSpringboardCareersGuidetoDataScience,wedidn’texpectthe
engagementit’dgarner.Thousandsofpeoplesignedupinafewdays,confirmingourbeliefthat
therewasascarcityofgreatadviceonwhatisanexcitingbutnebulousfield.
Inspeakingwithmoreandmorepeople,wefoundonlyafewgreatresourcesthatexplainedhow
tobreakintoadatasciencecareer.Therewereindividualstoriesandcollectionsofinterview
questions,butwecouldn’tfindafullguidetocovereverythingaboutthedatascienceinterview
processfromhowtogetaninterviewinthefirstplacetohowtodealwithanyofferedpositions.
Iwantedaguidecollectingperspectivesfrompeopleonbothsidesofthetable.Iwantedtotalkto
recruiterswhorefercandidates,hiringmanagerswhotableoffers,andcandidateswhohad
successfullymadeitthroughthedatascienceinterviewtodemystifythedatascienceinterview
processwithinsightsfrompeoplewhohadpreviouslygonethroughtheprocess.Icoauthored
thisbookwithSriKanajanaseniordatascientistinNewYorkCityatamajor
investmentbank.
AtSpringboard,we’vetaughtthousandsofdatascienceaspirantsthroughourmentored
workshops.Webuiltlarge,engagedcommunitiesofmentorsandalumni,whichaffordusaunique
vantagepointtodeliverreallifeperspectivesonthedatascienceinterviewprocess.
Itwasdifficultcollectingeverythinghere,likeitwasdifficultformanyofthecandidateswhomade
itthroughtheprocess.Someoftheleadersindatascience,includingtheChiefDataScientistof
theUnitedStates,hadtogothroughsixmonthsofwaitingbeforetheygotanoffer!Most
companies’datascienceinterviewprocessesaredesignedtoweedoutallbutthemostdetermined
andskilledcandidates.Itcanseem,attimes,likeahurdlepreventinganysanejobseekerfrom
entering.Yet,whiletheinvestmentcanseemimmense,thereturncanbeevengreater.
www.springboard.com
ULTIMATE GUIDE TO DATA SCIENCE INTERVIEWS
4
Datasciencehasbeencalledthesexiestjobofthe21stcentury.Datascientistsdon’tjustmake
goodmoney;theydrivesignificantsocialimpactfrommappingworldpovertytostopping
pandemicsbeforetheyevenhappen.DatascientistsunearthedtheidentityofBanksy,andthey
masteredtheartofpredictingbasketballscoresinMarchMadness.Workingindatascienceisn’t
aboutjusthavingagoodsalaryandgoodworklifebalance;it’saboutsolvingbigproblemsthat
matter.
Wewrotethisguidebecausewewantedtoyoutogofrombeingcuriousaboutdatascienceto
activelytryingtogetajobinthefield.Wewantedtounearthwhatittakesforyoutomakeit
throughthedatascienceinterviewprocess.Wewrotethisguidebecausewewantyoutorock
yourdatascienceinterview.
 
www.springboard.com
ULTIMATE GUIDE TO DATA SCIENCE INTERVIEWS
5
What is Data Science?
Beforeyoulookfordatascienceinterviews,youshouldknowwhatthetermmeansandwhat
you’regettingyourselfinto.
DJPatil,thecurrentChiefDataScientistoftheUnitedStates,firstcoinedthetermdatascience.
Adecadeafteritwasfirstused,thetermremainscontested.Thereisalotofdebateamong
practitionersandacademicsaboutwhatdatasciencemeans,andwhetherornotit’sdifferentfrom
thedataanalyticscompanieshavealwaysused.Whenpeopletalkaboutbigdataandusing
machinelearningtosolvedataproblems,theyareventuringintoawholenewfieldwhoseterms
arebeingdefinedrightnow.
Differentcompanieshavedifferingdefinitionsofwhatdatasciencemeans.Individualhiring
managersmaydifferaboutexactlywhatthey’relookingfor;theywillhireandinterview
accordingly.
Thisconfusionmakesthedatascienceinterviewprocessdifficultforalotofcandidates.Data
sciencecanhavevastlydifferentdefinitionsdependingonwhatroleyou’reapplyingforandthe
companyyou’reinterviewingwith.
 
www.springboard.com
ULTIMATE GUIDE TO DATA SCIENCE INTERVIEWS
6
Different Roles within Data Science
Let’sgothroughasampledata
scienceprojecttoelaborateon
thedifferentrolesyou’llseein
datascience.Adatascience
teammightbeassignedtouse
deeplearningtoclassifyimages
likeYelp’steamdid.
Millionsofphotosareuploaded
onYelpeverysingleday,butit
canbehardtogetimagesyou
wantforeachrestaurant.
Sometimes,thephotos
uploadedareallofthesame
category–maybethey’reall
photosofthefoodorthe
outsideoftherestaurant.A
holisticevaluationofa
restaurantrequiresimagesofdifferentkinds.
Youcanusemachinelearningtoautomaticallycategorizewhichimagesfallintowhatcategory.
Computerscan,withthehelpofatrainingset,tellyouwhetherornotanimageistheoutsideof
therestaurantoroffood.
Datascientistscreatethemodeltohelpmachinescreatethosedistinctions.Theywouldbeableto
thinkthroughthetypesofdatatheyneed,frommanuallytaggedphotostokeywordsinimage
captions.Thistendstobeamoreseniorlevelrole,astheyoftenmanagedataproductsfrom
www.springboard.com
ULTIMATE GUIDE TO DATA SCIENCE INTERVIEWS
7
endtoendanddealwithallfacetsofdatascienceproblems,fromalgorithmselectionto
engineeringdesign.
Dataengineerscreatesystemstosourcealloftheimagedataandstoreit,aswellasimplement
someofthealgorithmsdeterminedbydatascientistsatscale.Thistendstobearoleforpeople
withstrongtechnicalchopsbutmightnotknowasmuchaboutthetheoryofthealgorithmsthey’re
implementingatscale.
Dataanalystsqueryandpresentthebusinessimplicationsofthechange.Diditpleaseusers?How
muchmoretrafficdidYelpgenerateduetotherecentchange?Thesearequestionsdataanalysts
wouldask.Then,theycommunicatetheinsightstheyfound.Thisroletendstobefilledbymore
entrylevelpeopleandpeopleinbusinessfacingroleslearningtoapplytheirinsightsona
technicalbasis.
Therearemoreroleswe’llcoverindetaillater.Fornow,youshouldknowthatthedatascience
interviewprocessforallthreeofthesegeneralrolescanbevastlydifferentfromoneanotherand
infact,theyoftenare!

 
www.springboard.com
ULTIMATE GUIDE TO DATA SCIENCE INTERVIEWS
8
How Different Companies Think About
Data Science
Notonlyaretheredifferentrolesindatascience,therearealsodifferentcompanieswithvastly
differentinterviewprocesses!
Ingeneral,theserolescanbesplitintofourroughcategories.
1Earlystagestartups(200employeesorfewer)lookingto
buildadataproduct
WelcometothebeatingheartlandofSiliconValley.Theearlystagestartupisaromanticnotion,
butoneseeingastaggeringamountofsuccessinarapidamountoftime.Ifyoujoinanearlystage
startup,bepreparedtowearalotofhatsandpotentiallytakeonallthreedatasciencerolesatthe
sametime.Youwillneverhavetheresourcesyouneedinfull,sobepreparedtobescrappyand
tough.
Thebarwillbeespeciallyhighifthestartupinquestiondealswithdataasitsproduct.Aplatform
optimizingotherpeople’sdataorappliesmachinelearningtodifferentdatasetswillhavemuch
higherstandardsforhowtheythinkaboutdatathancompaniestryingtolearnfromtheirown
data.Thecofounderswilllikelybepioneersinthefieldofdatascienceorhaveledlargescaledata
scienceteams.TheywillbelookingforAplayerswhohavesignificantexperienceinthefieldor
tonsofpotentialanddrive.Ifyoujoinanorganizationlikethis,bepreparedforthelearning
experienceofalifetime,andbepreparedtobeheldtothehigheststandardpossiblewhenitcomes
todatascience.
Examplesofthiscompanytype:Looker,ModeAnalytics,RJMetrics
Samplejobpostings:DataAnalyst(Looker),SeniorAnalyst(ModeAnalytics)
www.springboard.com
ULTIMATE GUIDE TO DATA SCIENCE INTERVIEWS
9
 
Sizeofthecompany:143associatedonLinkedIn(1150companysize)
Howtoreadthisjobdescription:Focusoncommunicationandscriptinglanguagesfor
queryingandvisualizingdataindicatesthisisabusinessfacingrolewhereinsightsmustbe
communicatedtorelevantteams.
2Earlystagestartups(200employeesorfewer)lookingto
takeadvantageoftheirdata
Thebarwillbelowerifastartupismerelylookingtotakeadvantageofitsdataratherthanselling
adataproducttoothercompanies,butsincethesmartuseofdataisessentialtothecompetitive
advantageofastartup,youshouldstillexpectarelativelyhighbar.
www.springboard.com
ULTIMATE GUIDE TO DATA SCIENCE INTERVIEWS
10
Startupsinthetechindustrycontainalotoftechnicaltalent,buttheyneedsomebodytobridge
thebusinessandtechteams,especiallyiftherearecommunicationissuesbetweenthedifferent
teamsonhowdataisused.Bepreparedtoworkhardforthecompanytoembracebeing
datadrivenatalllevels,andbepreparedtobetheonewhobringsinnewtoolsandprocessesfor
collectingandusingdataatalllevelsoftheorganization.
Workingforacompanythatdealswithitsowndatabutdoesn’tthinkaboutdataatscalemaybe
anuniquechallengeasyou’llbecalledupontoenforceandspreadadatadrivenculture
throughouttheorganization.Bepreparedtoexerciseyourleadershipandcommunicationskills.
Lastly,B2BstartupsandB2Cstartupsdifferentiateinthedatatheyget.B2Bstartupsare
businesstobusiness;theysellsoftwaredirectlytolargecompanies.ThinkSalesforce.B2C
startupscatertomanyindividualcustomers.ThinkAmazon.Whenyou’redealingwithB2B
startups,you’relikelygoingtobefacedwithdatachallengesthataresmallinvolumebuthighin
detailandfeatures;startupsthatselldirectlytobusinessesdon’thavemanycustomers,butthey
focusmaniacallyontheonestheydohavesinceeachindividualcustomerwillbringinlotsof
revenue.B2Cstartupswillhavemoredataproblemsdealingwithvolumeandscaleastheywill
havemanymorecustomers,butthefocusonindividualcustomerswillbedilutedtofocuson
groupsofthem.AB2Bstartupmaydealwith1,000customers,allofwhompay$1,000amonth.
AB2Cstartupmaydealwith100,000users,buteachusermayonlygenerate$1inrevenuea
month!
Befamiliarwiththecompanyyou’reapplyingforandtheuniquedatachallengesitfaces.Research
thoroughly,andmakesureyou’reonlyapplyingforcompaniesthatfityourpassionsandskills.
Examplesofthiscompanytype:Springboard,Branch,Rocksbox,Masterclass,Sprig
Samplejobpostings:LeadDataScientistatBranch,DataScientist(Research)atRocksbox,,
DataScientistatMasterclass
www.springboard.com
ULTIMATE GUIDE TO DATA SCIENCE INTERVIEWS
11
Sizeofthecompany:37associatedonLinkedIn(1150companysize)
Howtoreadthisjobdescription:Lookingforageneralistwhocandivedeeperandstill
communicatedifferentinsightsindicatesthisisadatascientistrolethatwillbeverybroadin
termsofskillsetsdemanded.Thisroleisgoingtobeproactiveandentrepreneurial.
3MidsizeandlargeFortune500companieswhoare
lookingtotakeadvantageoftheirdata
www.springboard.com
ULTIMATE GUIDE TO DATA SCIENCE INTERVIEWS
12
Thelargestcompaniesintheworldknowthattakingadvantageoftheirdataisatoppriority.Some
willhaveestablisheddatascienceteamsthatarewellfunded,robust,andfedwithlotsofdata.
Somewillhavestartupliketeamswithintheorganizationtohelpthemtranslatetheirdatainto
businessinsights.Therearealotofcompanieshiringdatascienceteamsuponrealizinghow
importantdataistoremainingcompetitive.Usethistoyouradvantage;itcanbeeasierpassing
thedatascienceinterviewforalarge,prestigiousbrand.
Whilealotofthesecompanieswillhaveestablishedcorporateculturesandbureaucraciesthat
makeithardertoinnovate,theywillalsohavedataonmillionsofpeople.Imagineprocessing
logisticsdataforWalmartyouwillhavemillionsofdatapoints,andyourinsightswillmakea
differenceinthelivesofmillionsofpeople.
Whilethesecompaniesarenottraditionallyseenastheonesbuildingcuttingedgedatascience
solutions,thereisstillalotofgoodworkavailableforthosewhowanttoworkonchallenging
datasetswithtalentedteammates.
Examplesofthiscompanytype:Walmart,JPMorgan,MorganStanley,CocaCola,Capital
One
Samplejobpostings:DataScientist,ModeleratMorganStanley,DataEngineeratCapitalOne 
www.springboard.com
ULTIMATE GUIDE TO DATA SCIENCE INTERVIEWS
13
www.springboard.com
ULTIMATE GUIDE TO DATA SCIENCE INTERVIEWS
14
Sizeofthecompany:~30,000associatedonLinkedIn(10,000+companysize)
Howtoreadthisjobdescription:FocusonBigDatatoolsindicatesthatthisisgoingtobea
fairlyspecializedrolethatlooksintohandlingtheimmenseamountsofdataCapitalOneis
holding.
4Largetechnologycompanieswithwellestablisheddata
teams
Largetechnologycompaniesareabreedinandofthemselves.They’rethecontinuationofthe
startupobsessionwithdata,exceptnowtheyhavescaledtoapointdealingwithmillionsofdata
pointsormore.ThinkoftheUbers,theAirbnbs,theFacebooks,andtheGooglesoftheworld.
Withlargetechnicalteamsledbysomeofthemostbrilliantmindsintheindustry,datascience
roleshereareheavilyspecialized,andyou’llworkoncuttingedgeproblemswithdatathat
requiresferociouslyinnovativethinking.
Comehereifyoucraveachallengeandifyouwanttolearnalotwithalotofdatapoints.The
upsideisn’tasgoodastheearlierstagestartups,butyou’llgetgoodperks,goodsalary,andgreat
teammatesandagreatCVjobdescriptionincaseyoueverwanttomoveon. 
Examplesofthiscompanytype:Facebook,Google,Airbnb
Samplejobpostings:DataScientist,Oculus,DataScientistAirbnbMachineLearning
www.springboard.com
ULTIMATE GUIDE TO DATA SCIENCE INTERVIEWS
15

Sizeofthecompany:~16,715associatedonLinkedIn(10,000+companysize)
Howtoreadthisjobdescription:Focusonmultifaceted,innovativeskillsetshowsthisis
goingtobeanopenendeddatasciencerolethatwillbeexpectedtothinkofnewprojectsandlead
themfromendtoend.
 
www.springboard.com
ULTIMATE GUIDE TO DATA SCIENCE INTERVIEWS
16
Industries that employ Data Scientists
Datasciencealsovariesdependingontheindustry.Industrieshavecertainareasofknowledge
specifictotheindustryitself,andtheyinvolvedifferenttypesofdata.Aschoolwillbefocusedon
differentmetricsthanabank.
Ifyouhappentohaveapassionforacertainindustry,makesureitcomesoffwithkeywordson
yourCVandLinkedIn.Demonstratingwhyyouloveacertainindustryanddeepknowledgeofthe
industryitselfpositivelydifferentiatesyouasacandidate.
ThethreelargesthiringindustriesfordatascienceinO’Reilly’ssurveyofthefieldaresoftware
companies,consultingcompanies,andbanking/financecompanies.Thosethreeindustriesalso
tendtopaythemostfordatascienceprofessionals.
Differentindustriesalsovaryinthetypesofrolestheyhirefor.Software,medicineand
telecommunicationscompaniestendtobethelargesthirersofdatascientists.Software,
aerospace,andinformationtechnologycompanieshiremoredataengineers.Lastly,dataanalysts
tendtobehiredbyhealthcarecompaniesandconsulting/bankingorganizations.
Beawareoftheindustryyourpotentialemployerisin,andinferwhattheirdatascienceneedsare.

Youhavetobeawareofthedifferentroles,companies,andindustrieswithindatascienceto
understandexactlyhowyourdatascienceinterviewprocesswillgo.
Tododatascience,youmustbeabletofindandprocesslargedatasets.You’lloftenneedto
understandanduseprogramming,math,andtechnicalcommunicationskills.You’llalsoneedto
tailoryourskillsetandhowyoupresentyourselftothedifferentrolesandhiringcompanieswithin
theworldofdatascience.
www.springboard.com
ULTIMATE GUIDE TO DATA SCIENCE INTERVIEWS
17
Mostimportantly,youneedtohaveasenseofdeterminationtounderstandtheworld
throughdataandnotbedeterredeasilybyobstacles.
Thedatascienceinterviewprocessisdesignedtotestforthoseskillsandresilience.Bepreparedto
bechallengedoneverydimension.
GettingaDataScienceInterview
Thefirststepinthedatascienceinterviewprocessisn’tdealingwiththeinterview;it’sfindingitin
thefirstplace,aprocessthatinandofitselfcantakemonthsofeffort!
Wesurveyedabouttwentypeopleaboutthehardestpartsofthedatascienceinterviewprocessas
partoftheresearchforthisbook.Theanswerwegotbackhadlittletodowiththetechnical
questionswethoughtwerethehardest.Whiletechnicalquestionsrankedsecondwith68%of
respondents
selectingitasoneofthehardestpartsoftheinterviewprocess,awhooping80%of
respondents
selectedgettingadatascienceinterview!
Literaturewasscarceoutthereabouthowtogetaninterview,especiallyforpeopletransitioning
fromdifferentcareers.Wediveddeeper,andlookedthroughreallifecasestudiesinadditionto
differentresourceswe’vecuratedforyou.
 
www.springboard.com
ULTIMATE GUIDE TO DATA SCIENCE INTERVIEWS
18
Nine Paths to a Data Science Interview
Wefoundtraditionalpathstojobinterviewsthatcouldworktoacertaindegreeindatascience.
Wealsofoundnew,proactiveapproaches,especiallywithemergingstartups,where
nontraditionaltacticscouldgetcandidatestotheforefrontofthehiringrace.
TraditionalPathstoJobInterviews
Ifitain’tbroke,don’tfixit.Whilealotofthenew,proactivetacticswediscusscanhavealotmore
efficacy,it’salwaysgoodtoknowthebasics.
1- Data Science Job Boards and Standard Job Applications
Youcansubmityourresumesandcoverletterstocompanycareerssites.Then,youcanwaitand
hope.We’renotsayingtoavoidthisroute,butitshouldn’tbetheoneyourelyon.
UseIndeedandCareerbuildertosearchfordifferentdatasciencepostings.Then,therearespecific
jobboardsforthedatasciencespace,suchastheKaggleJobsBoard. 
2- Work with a Recruiter
Youcancontactrecruiterswhocanhelpputyouintouchwiththerightemployers.Thereare
recruiterswhospecializeindatascienceandtechnologyspaces.Theyaregatekeeperstojobsnever
listedinpublicoutlets.AquicksearchonLinkedInfordatasciencerecruitersnearyouwillhelp
youfindthemostrelevantmatches.
3- Go to Job Fairs
Jobfairsindatasciencearefarandfewinbetween,thoughHarvardandStanforddohost
computersciencejobfairsthathaveplentyofdatasciencejobsfortheirstudents.You’rebetteroff
www.springboard.com
ULTIMATE GUIDE TO DATA SCIENCE INTERVIEWS
19
attendingeithereventsormeetupswiththelocaldatasciencecommunityratherthanlooking
aroundforyourtraditionaljobfair.
ProactivePathstoJobInterviews
We’vecoveredthetraditionalpathstojobinterviews,theoptionsthathavebeenthedefaultof
jobseeking.Thesedays,gettinganoffersometimesrequireshustleandgritoutsideof
triedandtruetactics.Startupsprovidealargenumberofnewdatasciencejobs.Theircultureand
hiringtacticstrickleduptolargecompaniesthatadecadeagowerejuststartupsaswell.Theresult
isanewhiringenvironmentwhereoftentimes,onehastobeproactivetoreachdecisionmakers
whohaveknownnothingbutgritwhentheybuilttheirowncompanies.
4- Attend or Organize a Data Science Event
Youneedtofindpeopleinterestedinthedatasciencecommunitytofindhiddenopportunitiesand
becomeproactiveatintegratingintothecommunity.Thereareseveraleventswhereyoucando
this,fromlargerconferencestosmallercommunitymeetups.
Conferences
StrataConference
TheStrataConferenceisabigdatascienceconferencethattakesplaceworldwideindifferent
cities.Speakerscomefromacademiaandprivateindustry;thethemesorientaroundcuttingedge
datasciencetrendsinaction.Theconferenceallowsyoutolearnthetechnologybehinddata
science,andthereareplentyofnetworkingevents.
KDD(KnowledgeDiscoveryinDataScience)
KDDorKnowledgeDiscoveryinDataScienceisanotherlargedatascienceconference.It’salsoan
organizationthatseekstoleaddiscussionandteachingofthesciencebehinddatascience.
Membershipandattendanceattheseconferencesoffersamarvelouswaytocontributetogrowing
trendsindatascience.
www.springboard.com
ULTIMATE GUIDE TO DATA SCIENCE INTERVIEWS
20
NIPS(NeuralInformationProcessingSystems)
NIPS,orNeuralInformationProcessingSystems,isalargelyacademicdatascienceconference
focusedonevaluatingcuttingedgesciencepapersinthefield.Attendingwillgiveyouasneak
previewofwhatwillshapedatascienceinthefuture.
Meetups
We’velistedthemajorconferenceswherethedatasciencecommunityassembles,butthereare
oftensmallermeetupsthatservetoconnectthelocaldatasciencecommunity.
TheSanFranciscoBayAreatendstohavethemostdatameetups,thougheverymajorcityin
Americausuallyhasone.YoucanlookupdatasciencemeetupsnearyouwithMeetup.com.Some
ofthelargestdatasciencemeetups,withmorethan4,000members,areSFDataMining,Data
ScienceDC,DataScienceLondon,andtheBayAreaRUserGroup.
You’llwanttojointheevents,orcreateameetupyourselfifyoucannotfindanearbyevent.Our
directorofdatascienceeducation,Raj,gotajobbybecomingknownasadatascienceconnector.
HehostedalocalmeetupinAtlantaandinviteddistinguishedspeakersindatascience.Soon,he
wasknownasadatascienceinfluencer,andassoonastherewereopendatasciencepositions,he
wastappedtoapply.
5- Freelance and Build a Portfolio
SundeepPattemisadatainnovationleaderattheCaliforniaDepartmentofJustice.He’salso
mentoredforseveraldatasciencecourses,andasadatascientist,heworksoncreatingendtoend
solutionsthatextractvaluefromdata.Hehaspersonalwebsiteswithdifferentdatascience
projects.
Hisbreakthroughintodatasciencecamewhenhefoundanunsolvedprobleminenergy
sustainabilityandworkedtosolveit.Hewassoonapublishedauthorataprestigiousacademic
conference,andshortlythereafter,hewashiredtobecomeapracticingdatascientist.
www.springboard.com
ULTIMATE GUIDE TO DATA SCIENCE INTERVIEWS
21
Ifyou’reunsureofwhatdatayouwanttoanalyze,wehavealistof19free,opensourcepublic
datasetsyoucanexplore.
Ifyoufreelancearounddataproblemsyouloveandbuildincrediblesolutions,keeparecordof
everythingyoudoinanaccessibleportfoliothattellsstoriesaroundyourpassions.
6- Get Involved in Open Source and Open Data
Themostinterestingprojectsintheworlddon’tnecessarilyresideinsecretivecompanydatabases
anymore.TheyareofteninopensourcerepositoriesonGithub.ThisincludestheNatural
LanguageToolkitproject,whichhelpsdealwithhumanlanguageasadatasourceandthevarious
librariesthatmakeupthePythondatascienceandmachinelearningtoolkit.TheRcommunity
alsohostsmanyofitspackagesonaconsolidatedpublicwebsite.
ManyleadingCTOswillhirebasedonyourcontributiontoopensourceprojects,andmayeven
findyouthroughthatroute.It’seasytotellifsomebodyisabletoworkinateamandbuild
marvelousthingsthroughthetransparentglassofopensource.Makesureyoutakeadvantage.
7- Participate in Data Science Competitions
Ifthebroadconfinesofopensourceprojectsaren’tyourtypeofprojectsandyourcreativitythrives
bestinmoreconfinedsituations,considerjoiningadatasciencecompetition.
DatasciencecompetitionplatformslikeKaggle,DatakindandDatadrivenallowyoutoworkwith
realcorporateorsocialproblems.Byusingyourdatascienceskills,youcanshowyourabilityto
makeadifferenceandcreatethestrongestinterviewassetofall:ademonstratedbiastoaction. 
OneofourSpringboardmentors,SinanOzdemir,competedhiswaytoadatasciencejobbasedon
hisworkonproblemsonKaggle.Youcandothesame.
8- Ask People for Coffees, Do Informational Interviews
Attheendoftheday,yournetworkwillgetyouthebestchanceatanewjob.Youshouldseekto
knowmorepeopleinthefieldyouwanttoworkin,ifonlytogetanideaoftheproblemstheyhave
andwhichyoucansolve.
www.springboard.com
ULTIMATE GUIDE TO DATA SCIENCE INTERVIEWS
22
LegendaryentrepreneurandstrategistSteveBlankhasagreatframeworkforgettingcoffeeswith
peopletoobusytoseeyou,asmostdatascientistswillbe.Youhavetofindawaytoprovidevalue
ofsomekindandlooktogivethemafreshperspectiveontheproblemstheyface.
Thiscanculminateinaninformationalinterviewwhereyouseekadviceandinformationfrom
datascientistsinthefield.Ifyoudothisright,you’llconstantlygrowyournetworkofdatascience
opportunities,andyou’llunderstandmoreabouthowdatascienceworksinindustry.
9- Data Hackathons
Inlinewiththetrendofseeingworkinaction,datahackathonsofferyouanuniqueopportunityto
learndataskillswithamotivatedteam.Youwillhavetosolveadataprobleminacoupleofdays.
AnexampleofthiskindofhackathonistheDataWeekhackathoninSanFrancisco.Byteamingup
withotherstodeliverrealsolutions,you’lldifferentiateyourselffromotherjobcandidates.Many
employerslieinwaitathackathonsaswell,somecompaniesgoingasfarastosponsorhackathon
prizesinthehopesoffindingtheirnextdatascientist!
 
www.springboard.com
ULTIMATE GUIDE TO DATA SCIENCE INTERVIEWS
23
Working with Recruiters
ForthissectionweworkedwithAndyMusick,anAtlantabasedrecruiter:contacthimat
andy.musick@hotmail.comifyouwerelookingforanAtlantaareajob.Wealsoworkedwith
AnnaMeyer,adatasciencerecruiteratRobertWalters,arecruitmentagencyspecializedin
datascience.Feelfreetocontactheratanna.meyer@robertwalters.com.
HowtoApply
CV vs LinkedIn
Alotofpeopleouttheremayhaveatraditionalviewonwhatmakesforagoodjobapplication.
They’realreadymissingalargerpoint:thetraditionalviewisout.
Thereisafundamentaldifferencebetweenacademiaandworkinginanindustry,and
itstartsinhowyoupresentyourself.
Wetalkedwithrecruiters,students,andhiringmanagers,andtheyallagreedthatLinkedInwas
thegoldenstandardofrecruitment.HavingawelloptimizedLinkedInprofileallowsemployersto
sizeyouupandrecruiterstofindyoutherightopportunity.
Ifyou’renotmakingsureyoushineonLinkedIn,you’realreadylosingouttocandidateswhoare.
Whilearesumemayberequiredtogothroughtheprocess,itisn’tthemaindrawthatwillgetyou
inthedooranymore.Recruiterswillonlylookthroughyourresumeonceit’spresentedinfrontof
them,whileagreatLinkedIncouldleadtoinboundworkopportunitiesonaconstantbasis.
Unlikeinacademia,whereanimpressivearrayofpapersandacademicworkwillwinover
everythingelse,applyingforindustryjobsinvolvesbeingassuccinctaspossibleandlistingthe
www.springboard.com
ULTIMATE GUIDE TO DATA SCIENCE INTERVIEWS
24
impactyoudrivewithyouraccomplishments.resumesarenotsomuchreadasscanned.Keepthat
inmindifyou’regoingtobuildone.Arecruiterspendsanaverageofthirtysecondsonaresume.
KeyAdviceonResumes
1) Keepthemshort,preferablyunderapage.Rememberthatpeoplearescanningyourresume
forsignsofinterestbeforetheyeverconsiderdoingadeepdive.
2) Makesureyourskillsstandoutandarehighlighted(considerboldingrelevantskills).
Recruitersandhiringmanagerswilllooktoseeifyou’reatechnicalfitbeforelooking
further.
3) Haveclearjobheadings,andatmost,threeonelinepointsineachoneofyourjob
descriptions.Youwanttoclearlymarkhowyourexperiencetiesinwiththejob
requirementsyou’veappliedfor.
4) Demonstrateyourimpactwithnumbers!Don’tsaythatyou“didX.”Tellthehiring
managerwhateffectsXhad.Youwanttosayyoudiscoveredsomethingthathelped
thousandsofpeoplesavehoursoftimenotthatyousimplydiscoveredsomething.Write
“createdanautomatedsalesemailsoftwarethatgenerated$400,000”not“createdan
automatedsalesemailsoftware.”
KeyAdviceonLinkedIn
1) Don’tbeshy.Filloutasmanydetailsasyoucan;itmakesadifference.Mosthiring
managerswillwanttoseeyourLinkedInbeforetheyeverinterviewyou.
2) Makesureyourjobtitlesareclearandconsistentwithsearchtermsthatrecruiterswould
use.Sayingthatyouworkedasadatascientistorasadataanalystispreferredtocomingup
withyourownjobtitle.
3) Onewayyoucandifferentiatefromothersisaddingsomepersonalflavortoyourprofile.
Addsomeofyourinterestsandpassions,andmakesuretheyareevidentinyourLinkedIn.
Hiringmanagerslikeevaluatingcandidatesfortechnicalskillsandculturalfit.Beingableto
showthatyouhaveyourownuniquetakeontheworldwillonlyaddvaluetoyourjob
searchandhelpyoustandout.
www.springboard.com
ULTIMATE GUIDE TO DATA SCIENCE INTERVIEWS
25
4) Whileyoumightnotwanttotailoryourprofileforcertainjobsorindustries,makesureyou
knowwhatyou’relookingfor,andmakesurethatcomesoutonyourLinkedIn.Youwantto
beverydeliberateatconstructingyourprofilesothatitgetsyouthepositionyouwant.
Avoidlistingdataentryifyoudon’twanttogetentryleveloffers.Mentionspecific
industriesifyourheartissetonworkingonaparticulartypeofproblem.
Makesureyouknowwhatrolesyou’reapplyingfor,andapplyindustrykeywordsandskill
keywordsthatmatch.Interestedinadatasciencejobinfinance?Don’thesitatetoputindustry
terminologyalloveryourCVandLinkedIn.Ifyouhaveaskillthatyouresearchedisindemandfor
theroleyou’relookingfor,additliberally!Youcanresearchwhattechnologiesacompanyuses;
companieslikeYelpandAirBnBwilloftenblogabouttheirdataprojects.Ifyouseethattherolein
questiondemandsPythonandRskills,makesurethatyourCVandLinkedInmarksthoseskills.
EndorsementsalsoplayapositiveroleinthisregardwhenitcomestoLinkedIn,sodon’tbeshyat
askingpeoplewhohaveworkedwithyoutoendorseyourskillsandgiverecommendations.
MorerecruitersandhiringmanagerslookthroughLinkedInthanresumestoday.Arecruiterwill
lookataCVforanaverageof30secondsbeforediscardingit.Makesuretheimpactyou’vedriven
isfleshedoutwithstrongactionverbs,you’veformattedyourresumeandLinkedIntostandout,
andyou’vefilledthemwiththerightkeywords.
KeepinmindthatthisisafirststepandthatapplyingwithjustaCVorLinkedInwillgetyou
consideredatmostplaces,butnotwithanyparticularenthusiasm.You’llhavejoinedthequeueof
thousandsofothersapplyingthesameway,andyou’llprobablyneedtodomoretogetyourdream
job.Regardless,makesureyouoptimizeeverystepofyourapplication,includingtheCVor
LinkedInthatemployerswillinevitablylookover. 
CoverLettervsEmail
Acoverletterwasalwaysthestandardforacademicadvancement.Nowadays,therecruiterswe
talkedtoconfirmedthatcompaniesseldomreadthem.Ifyouwanttodifferentiatewhoyouare,
you’llhavetodoitonyourCVoryourLinkedIn.
www.springboard.com
ULTIMATE GUIDE TO DATA SCIENCE INTERVIEWS
26
Ifyou’regoingtobeproactive,sendabriefsummaryofwhatyou’vedoneinanemailtoahiring
manager.Thisservesasmoreofanexplainertheycansharewithotherpeopleinthecompany.
You’llwanttokeepitbriefnomorethanafewparagraphsatbestandyou’llwanttokeepthis
emailfocusedonthetopthreepointsthatdefinetheimpactyou’vedriven.
HowtogetReferencesandYourNetworktoWorkforYou
Mostpeopledon’trealizehowcriticalitistobuildandmaintainyournetworktogetyourfeetin
thedoorwiththedatascienceinterviewprocess.Thestrongestsignalhiringcompanieslookforis
strongreferrals,especiallyfrominternalsources.Ifyouhavesomebodyadvocatingforyouinside
theorganizationyou’reapplyingfor,thatcanensurethatyourCVwillbelookedover,anditcan
evengetyoutoskipstepsintheinterviewprocess!
Weinformallysurveyedsomeofouralumsgoingthroughthehiringprocess.Itturnsoutthata
referralfromaninsiderwithinthecompanyledtoa85%chanceofgettinganinterviewwiththat
particularapplication,whilethosewhoreachedoutcoldandonlyappliedwiththeirCVor
LinkedInorthroughthestandardformatonlyhadarounda10%chanceofgettinganinterview.
Pursuingtheformercanimproveyourjobhuntingprocessbyanorderofmagnitude.Ouralums
alsosaidthatthereferraldoesn’tevenhavetocomefromafriend,thefactthatanapplicationhas
beenreferredbyanexistingemployeeoftenguaranteesatleastaphoneinterview
.
Takealongtermviewonthisbyaddingvaluetodifferentpeopleinyournetwork,whether
that’sbeinggenerouswithadviceoncethat’saskedofyouorbeinggenerouswithintroductionsto
otherpeopleinyournetwork.Hopefully,bythetimeyou’relookingforajob,you’llhavebuiltupa
strongnetworkofpeoplealsointerestedindatasciencethatcanmaketherightintroductionsand
giveyoutherightreferrals.
Ifthatisn’tthecase,andyou’relookingtogetthosereferralsrightnow,youcanusewhatiscalled
theinformationalinterviewtechnique.Thisentailsreachingouttopeoplewhoareworkinginthe
fieldtogetasenseofwhat’sgoingonandwhattheirproblemsare.People,evencomplete
strangers,canbeverygenerouswiththeirtimeifyoushowthatyou’regenuinelyinterestedin
whatthey’redoingandyouoffertohelpaswell.
www.springboard.com
ULTIMATE GUIDE TO DATA SCIENCE INTERVIEWS
27
Lookforpeopleatmeetups,orspecificallytargetpeopleonnetworkssuchasLinkedIn,Angellist
andFounderDating.Presentyourintentionshonestly,butindicatethatyou’reveryinterestedin
thecompanyanddatascienceingeneral.Askforacoffeewhereyoucanaddperspectivetoa
problemthey’resolvingorlearnabouttheircompany.
Asamplescriptmightgoasfollows(whereyoucanaddsomebodyonLinkedInasafriendor
messagethemdirectlyonFounderDatingorAngellist):
Hi[name],
IwassuperinterestedintheproblemsAirbnbisfacingindatascience.I’vebeenaspiringto
breakintothefield,andbeingapassionatefollowerofthe
AirbnbNerds
blog,Inoticedthat
buildingtrustwithdata
isanimportantpartofwhatdrivesAirbnb.Basedonmybackgroundin
psychologyandstatistics,Imightbeabletohelpcomeupwithsomecreativeideasonhowto
fostertrust.
I’dlovetotakeyououttocoffeeandgetagreatersenseofwhatproblemsAirbnbhasperhapsI
canhelp!Wouldyouhavesometimeinthecomingweeks?
Cheers,
[yourname]
LinkstoyourLinkedIn,resume,portfolioand/orarecentproject
Ifyoureachouttoenoughpeopleandseekintroductionstopeoplethroughyournetwork,you’llbe
abletofindpeopleinanycompanytotalkwith.CheckoutyoursecondconnectionsonLinkedIn
andhowtheyareconnectedtoyou,whichyoucaneasilydothroughanyLinkedIncompanypage.
Here’sanexampleofacompanypageforAirbnb.
Onceyou’resetforaninformationalinterview,makesureyou’veresearchedthecompanyandthe
personyou’vetalkedwithbylookingatthecompanywebsiteandanyotherresourcesyoufind.
www.springboard.com
ULTIMATE GUIDE TO DATA SCIENCE INTERVIEWS
28
Youshouldhaveaprettygoodsenseofwhatproblemsthecompanyencountersonadaytoday
basis.
Theseinformationalinterviewsareagreatchancetoknowexactlywhatishappeningatacompany
andwhattheirprioritiesare,whichisgreatlybeneficialknowledgeinanactualjobinterview.If
youcomeinwellpreparedandpositionyourselfassomeonewhocanhelpthecompany,the
personyou’rehavingcoffeewithcouldbecomeastronginternaladvocateandhelpyoujump
throughtheusualrecruitinghoopstogetyourfirstroundinterview.
 
www.springboard.com
ULTIMATE GUIDE TO DATA SCIENCE INTERVIEWS
29
Preparing for the Interview
Hopefullyalltheworkyouputintogettingthedatascienceinterviewpaysoff,andyougetthe
emailthatsignifiesthestartoftheinterviewprocessforyou:acompanyrepresentativebeckoning
foraninitialphonecall.Here’swhatwillhappenandhowyoushouldprepare.
WhattoExpect
Thedatascienceinterviewisacomplexbeast,withbehavioralquestionsmixedwithabunchof
technicalquestions.You’vegottenprettyfarifyou’reabletogetaninterviewinthefirstplace,but
youstillhavefurthertogo.
Let’sstartfromthebeginningadatascienceinterviewwillbevastlydifferentdependingonthe
positionyou’reapplyingforandthehiringorganization.Certainorganizationswillbevery
rigorousandmakeyougothroughseveraltechnicalchallenges.Otherswilllookmoreatculturefit
and,especiallyifyouhavestrongreferences,getyoustraightthroughtothefinalround.
Themostrigorousprocesspossiblelookssomethinglikethis:
1- The Phone Screen
ThiswilltypicallybedonebysomebodyinHRandactsasafiltertosavehiringmanagerstime.
Sometimes,therewillbebasictechnicalquestionstoscreenoutcandidateswhoaregrossly
unqualified,butmostofthetime,thisphonescreeninvolvesestablishingthebeginningsofculture
fitandmakingsurethatthecandidatehasgoodenoughcommunicationskillstocomeoffwellin
theinterview.
Inthiscall,you’llwanttogetasenseofwhatproblemsthedatateamisfacingandthe
organizationalstructureoftheteamyou’reapplyingto.Comepreparedwiththoughtfulquestions
www.springboard.com
ULTIMATE GUIDE TO DATA SCIENCE INTERVIEWS
30
thatdemonstrateadeepunderstandingofthebusinessandthespacetheyoperatein,andbe
preparedtoaskthemattheend.
2- Take-home Assignment
Afterthephonescreen,companiesoftensendapreparedassignmentforcandidates,withsome
timepressurebeingapplied.Thisisagoodwaytoscreenoutcandidateswhomaybetechnically
weak,orwhomaynotbecommittedenoughtoinvestalotoftimeintherecruitmentprocess.
Somecompaniesdispenseofthisaltogether,butthosethatdoembracethetakehomeassignment
oftenuseitasatestingbartosavetheirhiringmanagerstime.
Anexampleofatakehomeassignmentisdoingadeepanalysisonaspecificdatasetprovidedfor
you.Whentheassignmentisdesignedwell,theassignmentisalsoanopportunitytolearnmore
aboutthetypesofproblemsyouwouldworkonifyouweretogetthejob.Here,you’dbeexpected
tostorytellaroundinsightsyou’dfindinthedata.Anotherexamplewouldbehavingadatasetwith
significanterrorsinitthatyou’dbeexpectedtoclean.Afinalexamplewouldinvolveworkingwith
aspecificproblemrelevanttothebusiness,suchasbuildingajobrecommendationsystemfor
applicantsbasedondatafromjobdescriptions.
Onlythosethatpassthebarofhavinggoodassignmentswilltalktoahiringmanagerfacetoface.
You’llgetweededoutquicklyifyourefusetodoitalltogether.
Takethetimetodotheassignment,andtrytoseehowitrelatestowhatproblemsthecompanyis
undergoing.Usingtheassignmentasawaytoseewhatkindofskillsyou’llbetestedonandhow
thecompanyinquestionisthinkingaboutyourroleensuresthatyoumaximizeyourtime.Thisis
whereyoucanshineinahiringprocessandshowhowyouaredifferentfromothercandidates.
3- Phone Call with a Hiring Manager
www.springboard.com
ULTIMATE GUIDE TO DATA SCIENCE INTERVIEWS
31
Youmayreceiveanotherphonecallscreenthatwillbefocusedoneithermathematicsand
statisticsquestionsorcodingquestions.Thiswillbedonebyahiringmanageroratechnical
person.Thiswilllikelybethefinalevaluationbeforeacompanyinvitesyoutoanonsiteinterview.
Thephonecallwilltypicallybesplitintothreecomponents.Sometimes,thisisdoneinonelong
call;othertimes,itisdoneinthreeshortphonecallsofaboutthirtyminuteseach.
Mathematical/StatisticalPhoneCall
You’llbeevaluatedoncoremathematicalandstatisticalconceptshere,whichwilldepend
somewhatonwhatroleandwhatcompanyyou’reapplyingfor.Webcompanieswilltendtofocus
onyourknowledgeofA/Bsplittesting,yourunderstandingofhowpvaluesarecalculated,and
whatstatisticalsignificancemeans.Energycompaniesmighttestyoumoreheavilyonregression
andlinearalgebra.Nomatterwhattypeofintervieweryou’retalkingwith,you’llwanttosketch
outtheentirethoughtprocessbehindyourproblemsolving.
Ifyou’reaskedaboutA/Bsplittests,describetheA/Bsplittestprocessindetail,fleshingoutwhat
pitfallstowatchoutforandleaningonanyexperienceyoumighthaveinthefield.Treatthe
questionlikeamathematicalproofandatestofyourabilitytostatisticallyreason,butdon’t
hesitatetoturnyourfinereyestodetailandacoherentstoryaboutwhythismatterstothe
companyathand.
CodingPhoneCall
Thispartoftheinterviewprocessisfairlytypicalandisalsotheclosesttoothertechnical
interviews.You’llbeevaluatedonyourability,overthephone,tosolvecodingchallengesby
presentingeitherpseudocode,orinharderinterviews,compilereadycode.Ifyou’reapplyingfora
dataanalystposition,thiswillswingmoretoaskingyouhowyou’dthinkaboutqueryingdatawith
SQL.Otherwise,you’llbeaskedquestionsintheprogrammingandscriptinglanguagesyou’ve
claimedexperiencein,fromJavatoPython.
www.springboard.com
ULTIMATE GUIDE TO DATA SCIENCE INTERVIEWS
32
YourinterviewermayalsousetoolslikeHackerRankorCollabedittoevaluateyouliveonline.In
thiscase,yourhiringmanagerwillwatchyouasyoutypeoutyoursolution:bereadyfor
approachessuchasthis,andtrainwiththosetoolsifyoucan!
Thereareplentyofgreatresourcesoutthereforcodinginterviews,fromCrackingtheCoding
InterviewtoInterviewCake.Usethemtoyouradvantage.
Practicemakesperfecthere.Makesureyouhaveacomfortablespaceandnaturalenvironmentfor
youtocode.Bepreparedtojotdowncodeonapaperandexplainitonaphonecall,orbeprepared
totypeinthecodeonalaptop.
Youwilloftenbeaskedaboutdatastructuresmorethananythingelse.Knowhashmaps,trees,
stacks,andqueuesverywell.Prepareforthisphonecalllikehowsoftwareengineerswould
prepareforacodinginterview,andyou’llpasswithflyingcolors.
CallwiththeHiringManager
Finally,you’llbepatchedthroughtothehiringmanager,whoisnowevaluatingyouonhowwell
youcommunicateandifyou’dfitwellontheteam.Thismaybeonaseparatecallfromthe
technicalphonescreens,oritcanbethelastpartofamegacallthatencompassesallthree.Inthis
call,thehiringmanageristryingtogetafeelforyourcharacter,yourmotivation,yourfitwith
theirteam,andyourrawintelligence.Mosthiringmanagershaveamentalmodelforwhotheyare
lookingfor.Thecloseryoufittoit,themorelikelyyouwillpasstoonsiteinterviews.
Thisiswhereyourworkwiththerecruiterbeforehandwillshine.Themoreyouknowaboutthe
problemsthehiringmanagerisfacingandthekindofpersonthey’relookingfor,thebetteryou’ll
bepreparedtopresentyourselfastheperfectfit.Tailoryourcommunicationstothatgoal,andbe
confidentandclear,andyou’llmakeittothenextround.Trytopassthe“airplane”testaswell;
imaginethehiringmanagerevaluatingwhetherthey’dliketospendhoursoftimewithyou.The
workplacewillforceyoutoworktogethercloselyandspendalotoftimetogether.Makesureyou
showthatyoucangetalongwithyourmanager!
www.springboard.com
ULTIMATE GUIDE TO DATA SCIENCE INTERVIEWS
33
4- On-site Interview with a Hiring Manager
Finally,ifyou’vemadeitthroughtheearliercalls,you’llmeetyourhiringmanagerfacetoface.
They’llbeevaluatingyoufrombothatechnicalandnontechnicalperspective.They’relookingto
ascertainifyou’reafit,andtheymaytestyouonyourtechnicalchopsbyhavingyouwhiteboard
differentscenarios.
5- Technical Challenge
Ifthisdoesn’thappentoyouduringtheonsiteinterview,preparetobechallengedonyour
technicalskillsinoneformoranother,especiallyforrolesthatleanmoretowardsdata
engineering.You’lloftenfindthatthisissimilartoasoftwareengineeringinterviewwhereyouwill
beaskedtowhiteboardandwritedownhowyou’dimplementcertainalgorithmsorsolvecertain
problems. 
Hereiswherestrongknowledgeofsoftwareengineeringconceptssuchastimecomplexity/BigO
notationandastronggraspofthemathematicsandstatisticsbehinddataalgorithmscantruly
shine.
6- Interview with an Executive
Ifyoupassthebarforyourhiringmanager,you’lloftendoafinalinterviewwithasenior
executive.Inastartup,thiswilloftenbethecofounderortheCEOthemselves.
Ifyou’vemadeitthisfar,congratulations!Don’ttakeitforgranted,butthisisasignthata
companyisleaningtoanofferforyou.Normally,onlycandidateswhohavepassedthetechnical
barwillgethere,sonowyouneedtoemphasizeexactlyhowyoucandriveimpactwithyour
knowledgeofthebusinessitself,andthespecificproblemsitfaces.Atthispoint,you’renot
lookingtoproveyourselfsomuchastoavoidglaringerrors.
www.springboard.com
ULTIMATE GUIDE TO DATA SCIENCE INTERVIEWS
34
What a data scientist is being evaluated on
Position Title
Mathemat
ics/Statist
ics (e.g.
P-value
analysis,
AB
testing)
Databas
e
Querying
(SQL)
Algorithm
s (e.g.
Supervise
d learning,
Entity
Resolutio
n)
Software
Engineerin
g (e.g.
Python,
Java,
Object
Oriented)
Big
Data/Systems
Engineering
1
(e.g. Spark,
HBase, Hadoop)
Product Data Scientists
2
Medium
Medium
Medium
High
High
Data Engineering
Low
Medium
Low
High
High
Data Scientist
High
Medium
High
Low
Low
Business Intelligence Data
Scientists
Medium
High
Medium
Low
Low
Data Analyst
Low
High
Low
Low
Low
Differentdatascienceroleswillhavevastlydifferentexpectationsondifferentskillsets.Whilea
dataengineermightnotbeexpectedtohavemanybusinesspresentationskills,theyareexpected
todominatealltypesofprogrammingchallenges.Conversely,adataanalystwillleanmoreon
theirSQLskillsandnotbeexposedtoheavytechnicalproblems,buttheywillbeexpectedtobe
topnotchpresenters.
Thistableimpliestheindustrydemandanddifficultyofthepositionsfromtoptobottom,with
ProductDataScientistsbeingthemostindemandfortheirspecialized,difficulttoacquireskills.
Knowwhatroleyou’reapplyingfor.Seektoscoutoutexactlywhatneedsacompanyislookingfor
andwhatroletheyaretryingtofityouin;itwillhelpyounavigateandpredicttheirdatascience
interviewprocess.
1Thisismoreinlinewithdealingwithsettinguplargescaledataengineeringplatformsandintegratingvarioustechnologies
together.
2Thesedatascientiststypicallybuildthealgorithmandproductionizeitthroughthedataengineeringinfrastructure.E.g.
Theywouldbuildtherecommendationsystemalgorithmandproductionizetherecommendationsystemliveonthe
platform.
www.springboard.com
ULTIMATE GUIDE TO DATA SCIENCE INTERVIEWS
35
Here’sahighleveloverviewofthespecificroles:
ProductDataScientist:Endtoenddatascientistwithdataengineeringskills.Productdata
scientistsleadteamstobuildadataproduct.Theytweakalgorithmsandhaveastrongsayinhow
thedataisservedtoendusers.Theywilloftenhavetheengineeringabilitytodeliveronthose
ideas.
DataScientist:Theunicornmixoftechnicalskills,businessskills,andmathematicalknowledge.
Adatascientistunderstandshowtocreateandoptimizedataalgorithms,andhowtoexplaintheir
findings.Theymayneedtoknowlessprogrammingthantheirdataengineerpeers,butthey’ll
neverthelessneedtounderstandenoughtodealwithdataatscale.
BusinessIntelligenceDataScientists:BusinessIntelligenceDataScientistsarefocusedon
gettingbusinessinsightsoutofdata.Theywillunderstandenoughaboutstatisticalmethodsand
differentmachinelearningalgorithmstodifferentiatethemselvesfromdataanalysts.Theybuild
dashboardsandcompletevariousanalyticalstudiestohelpthevariousteamsmakebetter
decisions.
DataEngineering:Adataengineerisn’toftencountedontohaveadvancedknowledgeofthe
statisticsandmathematics,buttheywillhavetoaceeverytechnicalchallengeouttheretoprove
theycandealwithimplementingalgorithmsonmassiveamountsofdata.
DataAnalyst:Anentrylevelrolethatreliesheavilyonmakingoneoffreportsbylooking
throughdataandinterpretingtheresults.ThisroletypicallyrequiresastrongknowledgeofSQL
andExcel.
The Categories of Data Science Questions
Behavioral Questions
www.springboard.com
ULTIMATE GUIDE TO DATA SCIENCE INTERVIEWS
36
Thedatascienceinterviewprocessinvolvesalotofbehavioralquestions,similartoanyother
interview.Theinterviewerintendstotestforyoursoftskillsandseeifyoufitinculturallywiththe
company.
1. Tellmeaboutadatascienceprojectyouhavedoneinthepast?
Intent:Theintentofthequestionistounderstandthedepthofknowledgeand
contributionsyouhavefromyourpastexperiences.Ittestsyourabilitytotellastory
aroundyourworkandwhetheryoucantieittoimpactonthecompanyyouworkedwith.
HowtoAnswertheQuestion:
■Trytodescribeaprojectthatdemonstratesbothproductandengineering
experience,i.e.theprojectprovidedtheanalyticalinsightandproductionised
theinsighttomakeitactionable.Forexample,ifyouidentifiedkeytopicsina
textdatasetthroughtopicextractiontechniques,youshouldexplainhow
thesetopicsfurtheredcompanygrowthinadataproduct.
■Gointodetailaboutyourspecificcontributionandtheoutcomefroma
businessgoalperspective.Theinterviewerwantstoknowwhatyou
specificallydidwhiletryingtounderstandtheoverallgoaloftheproject.
■Rehearseyourexperiencesmanytimes.Thisisaverycommonquestion,so
have23gotoprojectsyoucangointoextremedetailabouteloquently.
2. Whathaveyoulikedanddislikedaboutyourpreviousposition?
Intent:Theintentofthequestionistoidentifywhethertheroleyou’reinterviewingforis
suitableforyou,andtoidentifywhyyou’removingonfromapreviousposition.
HowtoAnswertheQuestion:
■Understandtherolewell.UsetheHRcontacttogetinsiderinformationabout
theroleanditschallenges.TheHRpersoncanbeatreasuretroveof
informationabouttherole,team,history,andkeyimmediatebusinessgoals.
www.springboard.com
ULTIMATE GUIDE TO DATA SCIENCE INTERVIEWS
37
■Avoidtalkingaboutissueswithspecificpeople,andbeprofessionalwhen
talkingaboutwhatyoudisliked.Introspectcarefullyandtalktowhatmakes
youpassionate.Forexample,talkaboutderivinginsightsfromdataand
conveyingthemtomanagementinanactionablewayassomethingyouenjoy.
Youcouldalsotalkaboutlearningnewtechnologiesthatmakedatascience
moreactionablethroughtheorganization.Youcoulddislikehowthe
organizationisnotplacingdatascienceatthecenterofitsstrategyorthatthe
companyhashadsignificantattritionattopmanagementlevelandthe
directionoftheteamisunclear.Keepitpositive,pointsoriented,andaway
frompersonalsituations.
● Bad:Ihatedthatdatascientistswerealwaysputbelowtheengineers
andthatmanagementdidn’thaveacluewhatthecompanydirection
was!
● Good:IrealizedIwantedtoworkinacompanywheredatascienceis
partofitscorestrategyandthecompanyhasacleardirection.
3. Tellmeaboutasituationinthepastwhereyouhadtoconvinceothersaboutyourposition
onaspecificmatter.Whatwastheoutcome?
Intent:Theintentistofindouthowgoodareyouatdefendingyourpositionandyour
abilitytoengenderchangewithinateam.
HowtoAnswertheQuestion:Trytofindanexamplewhereyouweresuccessfulat
makingthechangeandthatthechangeisquantifiableinitsimpact.Ifpossible,useadata
sciencetypeexampleifyouhaveone.It’simportantthatyoudemonstrateyour
communicationandleadershipskillshere.
Mathematics Questions
Questionsaboutthemathematicsanglewillcomefordatascientistroleswhereyouareexpected
notonlytoimplementalgorithms,butalsotweakthemforspecificpurposes.
www.springboard.com
ULTIMATE GUIDE TO DATA SCIENCE INTERVIEWS
38
1HowdoestheLinearRegressionalgorithmfigureoutwhatarethebestcoefficientvalues?
(ThiswasaquestionaskedinC3Energy’sDataScientistinterview)
Rationale:Theintentofthequestionistoseehowdeeplyyouunderstandlinearregression,
whichiscriticalbecauseinmanydatascienceroles,youwon’tjustworkwithalgorithmsinablack
box;you’llimplementtheminsomeway.Thiscategoryofquestion(andyoucouldseeitfromany
typeofalgorithm)testshowmuchyouknowaboutwhatisactuallyhappeningbeyondthesurface.
HowtoAnswertheQuestion:Traceouteverystepofyourthinkingandwritedownthe
equations.Describeyourthoughtprocessasyou’rewritingoutthesolution.
TheAnswer:Atthehighestlevel,thecoefficientsareafunctionofminimizingthesumofsquare
oftheresiduals.Next,writedowntheseequationswhilepayingcarefulattentiontowhatisa
residual.Togofurther,considerthefollowing:
1. Writetheminimizationgoal(ideallyinlinearalgebraic(matrix)notation)ofminimizingthe
sumofsquaresoftheresidualsgivenalinearregressionmodel..
2. Solvetheminimizationequationbyillustratingthatthesumofsquareoftheresidualsisa
convexfunction,whichcanbedifferentiatedandthecoefficientscanbederivedbysetting
thedifferentiationto0andsolvingthatequation.
3. Describethatthecomplexityofsolvingthelinearalgebrabasedsolutionin#2isof
polynomialtimeandamorecommonsolutionisbyobservingthattheequationisconvex
andhencenumericalalgorithmssuchasgradientdescentmaybemuchmoreefficient.
StatisticsQuestions
Agraspofstatisticsisimportantforsolvingdifferentdatascienceproblems.You’llbetestedon
yourabilitytoreasonstatisticallyandyourknowledgeofthetheoryofstatistics.Bepreparedto
reciteyourknowledgeaboutstatisticalconceptslikeTypeIerrorandTypeIIerrorflawlessly,and
bepreparedtodemonstrateyourgraspofdifferentprobabilitydistributions.
1WhatisthedifferencebetweenTypeIerrorandTypeIIerror?(OuralumnusNiraj
encounteredthisquestion).
www.springboard.com
ULTIMATE GUIDE TO DATA SCIENCE INTERVIEWS
39
Rationale:Companieswillwanttotestyourgraspofdifferentbasicstatisticalconceptstotest
howgoodyouarewiththefundamentalsofstatisticsandseehowyoucommunicatedifferentideas
youmaynotoftenapplywiththesometimestechnocraticlanguageembeddedinstatistics.
HowtoApproachyourAnswer:Benononsense,andcommunicateclearlywhateveryouare
askedtodefine.
TheAnswer:TypeIerroriswhatisreferredtoasa“falsepositive,”ortheincorrectrejectionof
thenullhypothesis.TypeIIerroriswhatisreferredtoasa“falsenegative,”ortheincorrect
acceptanceofthenullhypothesis.Youmaywanttocommunicateyourgraspoftheconceptswith
anexampleandhowitmightberelevanttothebusinessathand.TypeIerrororafalsepositive
wouldbetellingamantheywerepregnant,whileTypeIIerrorwouldbetellingapregnantwoman
theyweren’t.Ifyouwererunningafrauddetectionbusiness,youmighthaveaveryhightolerance
forfalsepositives(aclientwillnotfussaboutanemailonthepotentialoffraud),butafalse
negative(notdetectingfraudwhenitishappening)couldbedisastroustoyou.
2Thiswasaquestionforadatascientistpositionatabiginsurancecompany.Supposea
populationisdividedintotwogroups:aggressivedriversandnonaggressivedrivers.40%of
thepopulationareaggressivedriverswhile60%arenonaggressivedrivers.Theprobabilityof
anaggressivedrivergettinginto3accidentsinoneyearis15%.Theprobabilityofa
nonaggressivedrivergettinginto3accidentsinoneyearis5%.Johnisknowntohave3
accidentsinthepastyear.Whatistheprobabilitythatheis(a)anaggressivedriver,and(b)a
nonaggressivedriver?
Rationale:AlotofcompanieswilltestyourBayesianinferenceskillsasaprimerforhowyou
thinkstatistically.Bayesianprobabilitycontrastswithfrequentistinterpretationsofstatistics,and
yourabilitytoreasonthroughanyBayesianproblemwillshowyouhaveaquickgraspofstatistical
conceptsandthementalmathneededforit.Ifyouneedarefresher,oneofSpringboard’smentors
WillKurtrunsablogcalledCountBayesie,andhehasawonderfulguidetoBayesianstatistics.
HowtoApproachyourAnswer:
TheintentofthequestionistoseeyourlevelofunderstandingBayesianprobability.Sketchoutall
ofyourassumptionsandthecalculationsyou’redoingforyourinterviewerinalogicaland
organizedfashion.
www.springboard.com
ULTIMATE GUIDE TO DATA SCIENCE INTERVIEWS
40
TheAnswer:Writeoutwhatyouknow.
Probabilityofaggressivedriversinthepopulation=40%or0.4
Probabilityofnonaggressivedriversinthepopulation=60%or0.6
Probabilityofaggressivedriversgettingintothreeaccidentsayear=15%or0.15
Probabilityofnonaggressivedriversgettingintothreeaccidentsayear=5%or0.05
You’llwanttounderstandtheconceptofpriorsandposteriorsforBayesianequations.Aprioris
whatyouaregivenbeforetheproblem,datathatyoureceive.Theprobabilitythatsomebodyisan
aggressivedriverinthepopulationisapriorassumptiongiventoyouthatyoucannotchange.The
posterioristheprobabilityyouderivefromusingtheBayesRuleontheseassumptions(P(A/B)).
BayesRule
Thefirstquestionis“whatisthechanceJohnisanaggressivedriverifhe’sbeenin3accidentsa
year?”
Visually,you’rereallytryingtodrawaVenndiagramofprobabilities:ofallofthepeoplewhohave
beenin3accidentsayear,howmanyareaggressivedrivers?Howmanyarenot?
www.springboard.com
ULTIMATE GUIDE TO DATA SCIENCE INTERVIEWS
41
Thereisa67%probability(really66.66%repeating)thatsomebodywhogetsintoa
3accidentsayearisanaggressivedriver.Thisisnowyourposterior.
Theprobabilitythatsomebodywhogetsinto3accidentsayearisnonaggressiveis
justtheflipsideofthat.10.6666=0.33333repeating,or33%probability.
3Whatisprobabilitydistributiontype(orshowthederivationofthepdf)youwoulduseto
describethefollowingrandomvariables?
a. Probabilityofkcustomersarrivingtoarestaurantwithinadurationoftminutes
b. TheprobabilityoftheheightofapersoninacrowdbeingatleastXinches
c. Theprobabilityofthesumoftwo6sidedfairdicesbeingY
d. TheprobabilityofhavingkheadsthrownoutofNcointhrows
Rationale:Thisquestiontestsyourknowledgeofprobabilitydistributionsandtestswhetheror
notyouknowwhatmodelstousegivenhowyourdataisorganized.
HowtoAnswertheQuestion:Explainyourassumptionsaboutthedataandthedetailsofhow
thedistributioninquestionfitsthemodel.Beabletovisualizedistributionsandexplaintothe
interviewerwhythedistributionyouvisualizefollowsthemodel.
Answer:
www.springboard.com
ULTIMATE GUIDE TO DATA SCIENCE INTERVIEWS
42
a. Poissondistribution.Thisisassumingthatcustomerarrivalsareentirelyindependentfrom
eachother.
b. Normaldistribution.NotethatinacontinuousdistributionthelikelihoodofbeingexactlyX
inchesiszero.
www.springboard.com
ULTIMATE GUIDE TO DATA SCIENCE INTERVIEWS
43
c. P(sum(x1+x2)={0,1,(2,12),(3,11),(4,10)...36})={0,0,1/36,2/36,3/36,…}.Youcanplotthis
outwherethexaxisisthesum,andtheyaxisistheprobability.Illustratethatthisisa
probabilitymassfunctionvsacontinuousprobabilitydistributionfunction.
d. Binomialdistribution.P(kisthenumberofheadsinNthrows):
www.springboard.com
ULTIMATE GUIDE TO DATA SCIENCE INTERVIEWS
44
Notethatthisvisualizationsaysthereisa25%chanceyouwillget5coinsoutof10tobeheads. 
www.springboard.com
ULTIMATE GUIDE TO DATA SCIENCE INTERVIEWS
45
CodingQuestions
Alargepartofadatasciencerole,especiallyifitismorefocusedtodataengineering,is
programmingtoimplementalgorithmsatscale.Bepreparedtofacesomethingsimilartoa
softwareengineeringinterviewwhereyou’llbetestedonyourexperiencewiththetechnicaltoolsa
companyusesandyouroverallknowledgeofprogrammingtheory.
1SQLGivenatableoftransactions(Transaction_ID,Item_ID,quantity,purchase_date
(MM/DD/YY))andanothertableofprices(item_ID,price),givethefollowinginformation:
1. Totalrevenue
2. Totalnumber/average/standarddeviationofpurchasequantitiesforthesetofweekdays
(MondayFriday)orderedbydescendingnumberofpurchases.
3. Numberofitem_ID’sthatwereNOTpurchasedintheweekdays.
Exampletableoftransactions(definedastransactions):
Transaction_ID
Item_ID
Quantity
Purchase_Date
1
1
5
06/28/2016
2
2
3
06/27/2016
3
2
5
06/27/2016
4
2
1
06/26/2016
Exampletableofprices(definedasprices):
item_ID
Price
1
$2
2
$3
www.springboard.com
ULTIMATE GUIDE TO DATA SCIENCE INTERVIEWS
46
Rationale:TheuseofSQLtoquerydatabasesisprevalentinlargerstartupsandestablished
companieslookingtoleveragetheircompany.Ifyouareadataanalyst,yourtechnicalinterview
mayexclusivelybeSQLquestions.Understandinghowtogetdatatherightwaycanmakethe
differencebetweengettingajobandnot.
HowtoAnswertheQuestion:Youwilloftenbeaskedtosketchoutyourcodeonpaperor
workwithacollaborativecodingtoollikeHackerEarthwhereyouwillbecodingintheinterpreter
andyourcodeisseenlivebyyourinterviewer.Makesureyoutryforthemostefficientsolution
withasfewerrorsaspossiblegivenashorttimeconstraint.UsesomethinglikeSQLFiddleifyou
wanttopracticeyourSQLqueryingskills!
Answer:
1. SELECTsum(a.quantity*b.price)
FROMtransactionsASa
JOINpricesASbONa.item_ID=b.item_ID
Thiswilljointhepricecolumnfromthepricestableontothetransactionstable,allowingyouto
multiplythequantityofeachitemwithitspriceandthentosumupthatmultiplication.Thiswill
yieldananswerof$37forourtwoexampletables.
2. SELECTDAYOFWEEK(purchase_date),
sum(quantity),
avg(quantity),
std(quantity)
FROMtransactions
WHEREDAYOFWEEK(purchase_date)BETWEEN2AND6
GROUPBYDAYOFWEEK(purchase_date)
ORDERBY2DESC
ThisquerywillusetheDAYOFWEEKfunctioninMySQL,whichreturnsanumberindexofwhich
dayacalendardayis,andreturnsavaluefrom1and7,with1correspondingtoSunday,and7
www.springboard.com
ULTIMATE GUIDE TO DATA SCIENCE INTERVIEWS
47
correspondingtoSaturday.Filtering,selecting,andthenorderingbydescendingquantities
satisfiesthequestionoftable2. 
Ifyouranthequeryonthesampletable,you’dgetthefollowingoutput,with2correspondingto
Monday(June27th,2016):
3. Twoapproaches(usingLeftJoinvs.GroupBy):
a. SELECTCOUNT(DISTINCTA.item_ID)
FROMtransactionsA
LEFTJOIN
(SELECTpurchase_date
FROMtransactions
WHEREday_of_week(purchase_date)IN(Monday,
Tuesday,
Wednesday,
Thursday,
Friday))ASBON
A.Transaction_ID=B.Transaction_ID
WHEREB.purchase_date=NULL
b. SELECTCOUNT*
FROM
(SELECTitem_ID
FROMtransactions
WHEREIsWeekDay(purchase_date)!=TRUEgroupby
item_ID)
Eitherapproachwillnarrowdownatableofitemsthatwerenotpurchasedontheweekend,then
applyaspecialcounttoit.
TipsforSQLQuestions:
www.springboard.com
ULTIMATE GUIDE TO DATA SCIENCE INTERVIEWS
48
1. Dosmallqueriesfirstinsteadofgoingtothesubqueries.Breaktheproblemdownto
specificintermediatetables,anddothequeriesforthoseintermediatetablesfirst.
2. Becarefulofthecolumnyoudothejoinon.Askwhetheryouwanttokeeprowswhere
therewasn’tamatch(i.e.leftjoinifneeded).
3. Ifyoudon’tknowtheexacttransformationfunction,assumetheexistenceofone,statethe
input/outputtotheinterviewer,andmoveon.
2DevelopaKNearestNeighborsalgorithmfromscratch
(algorithmcoding)
Rationale:Showingyoucanwriteoutthethinkingbehindanalgorithmanddeployitefficiently
inagiventimeconstraintwillbeacriticalwaytoevaluatedataengineeringskills.Thiskindof
questionwillbeaskedofdatascientistswhohaveknowledgeofbothalgorithmsandtheir
technicalimplementation,ordataengineerswhoaregivencontextonwhatisthealgorithm.This
questionscanbeaskedofanyalgorithm,butmostofthetimeinterviewerswilluserKnearest
neighbours,asit’srelativelyeasytocomeupwithcodethatcanwork.
HowtoAnswertheQuestion:First,clarifythequestion.Givenafeaturevector,findthe
euclideandistancefromthatvectortoeveryotherknownvector,andtaketheclassthatisthe
majoritywithintheclosestKvectors.Thisparticularquestiontestsyourunderstandingofmatrix
computationandhowtodealwithvectorsandmatrices.Startbygoingthroughasamplesetof
inputsandoutputs,andmanuallyderivetheanswer.Also,keepaneyeonthetime/space
complexity.Inthesolutionbelow,eachpredictionisofO(2N+NlogN)timecomplexitywhereN
isthenumberofrowsoftrainingdata.
Youwillwanttowritedownyoursolution.Syntaxcounts,andsodovariousfaultsthatwillstop
yourcodefromcompilingproperly,butitdoesn’tcountasmuchasexpressingthelogicbehindthe
algorithm,andshowinghowyoucanapplyalgorithmicthinkingtotheplaneofcomputerscience.
Solution:
www.springboard.com
ULTIMATE GUIDE TO DATA SCIENCE INTERVIEWS
49
Othercodingquestionscanbemorebigdataspecific.Forexample,askingaboutmapreduceisa
typicalquestioninthecasethatthepositionrequiresanalysisofverylargedatasets.Questions
hereaskhowtotaketheaverageofalargedatasetorfindthemostfrequenteventinanevent
stream.
3HowdoeswordcountmapreduceworkonHadoop?
Rationale:YouwillgetquestionsaboutHadoopandbigdatatoolsifyouindicateonyourCV
thatyouhaveexperiencewiththem,orifthecompanyinquestiondealswithmassivedatasets.
LargerFortune500companiesandtechstartupsthathavescaledbeyondmillionsofusersare
likelytochallengeyouonyouruseofbigdata.Youshoulddemonstrateaknowledgeof
mapreduce,whichcancomefromworkexperienceorplayingaroundwithmassivedatasetson
yourown.HortonhasresourcesdedicatedtohelpingpeoplelearnMapReduceifyouneedto
brushup.
HowtoApproachtheAnswer:Thisquestionseeshowdeeplyyouunderstandthemapreduce
frameworkonHadoop.ThisistypicallydoneusingJava.Althoughthewordcountproblemisan
extremelycommonlyunderstoodone,knowinghowit'simplementedwithintheJavaHadoop
frameworkistheimportantpiecehere.
www.springboard.com
ULTIMATE GUIDE TO DATA SCIENCE INTERVIEWS
50
Answer:Thedrivercodewouldsetupthejobandconfiguration.IfthedatacomesfromHDFS
andoutputiswrittentoHDFS,addtheinput/outputpathtothejobtothosedirectories.Thenthe
mapperjobwouldtakeeachlineinthefileandemitavalueof1foreachwordasthekey.Notethat
thedatapassedbetweenmapperandreducermustusetheHadoopdatastructuressuchasText
andIntWritablessincethesearemoreefficientforbytearrayserializationvs.primitivetypessuch
asStringsandInts.Themapperoutputwouldthenbecollectedineachexecutor,andthenthe
combinertaskwouldbeexecuted.Thecombinerisalocalaggregatorthatisoptionallysetto
reducetheamountofdatasentbetweenthemappersandreducers.
Onceallthemappersarecomplete,onlythencantheshufflephasebegin.Youmightobserveyour
jobsstuckat33%reducer,whichimpliesthattheshufflephaseiswaitingonthemappersto
complete.Onceallthekeysaresenttothereducersbasedonthisshuffle,thesortphasebeginson
eachreducer.Afterthat,thereducelogicisexecuted,andtheoutputcanbewrittentoanother
HDFSfile.
Commonfollowupquestionsininterviewswouldbetoestimatethetimecomplexityofthis
algorithm,andtheamountofdatathesystemwritesorcommunicatesbetweenmachines.Don’t
forgettotakeredundancyintoaccount,i.e.aHadoopsystemusuallymakesmultiplecopiesofdata
incaseamachinegoesdown.
ScenarioQuestions
1Ifyouwereadatascientistatawebcompanythatsellsshoes,howwouldyoubuildasystem
thatrecommendsshoestoavisitor?(QuestionaskedinVerizonDataScientistInterview)
Rationale:Thisquestiontestshowyouthinkaboutyourworkintermsofdeliveringproducts
fromendtoend.Scenarioquestionsdon’ttestforknowledgeineveryfield;theyaresettoexplore
aproductfrombeginningtodeliveryandseewhatlimitsthecandidatewouldhave.Whilealso
evaluatingforholisticknowledgeofwhatittakestomanageateamtodeliverafinalproduct,this
questionistoseehowthecandidatewouldfitintoateamsituation.
www.springboard.com
ULTIMATE GUIDE TO DATA SCIENCE INTERVIEWS
51
Typically,datascientistswillbeaskedthisquestion,whiledataengineersoranalystsmightbe
askedforspecificpartsofthescenariorelevanttothem.Dataengineersmightbeaskedtothinkof
howtoimplementacertainalgorithmatscalewithouthavingtothinkofthealgorithmitself,while
dataanalystsmightbeaskedwhatdatathey’dquerytodetermineusers’historicalpreferencesfor
shoes.
HowtoAnswerthisQuestion:Beveryhonestastowhereyoucanaddalotofvalue
(emphasizewhatpartsyou’vehadexperiencein),butdon’tbeshyaboutwhereyouexpecttogeta
littlebitofhelp.Trytorelatehowyourtechnicalknowledgecanhelpwithbusinessoutcomes,and
alwaysenumeratethethoughtprocessbehindyourchoicesandtheassumptionsthatguidethem.
Don’thesitatetoaskquestionsthatcanbettertailorfityouranswer.
Answer:Breaktheanswertotwocomponents:DatascienceandDataengineering
Let'sdiscussthedatascienceelementfirst.Ifitisanewcompanythatdoesnothavemuch
historicaluserdata,gowithitemitemsimilarity.Ifthenumberofdifferentitems/shoesis
extremelylarge,considerusingmatrixfactorizationtechniquestoreducethedimensions.
Ifyouhavehistoricaldataarounduserpreferences(e.g.ratingsofshoes),youcanusea
collaborativefiltertypeapproach.Mentionspecificallytherowsandcolumnsofthematrixyou
generatewitheitherapproach.Thendiscusswhatkindofsimilaritymetricsyouwouldtry.E.g.
euclideandistance,Jaccardsimilarity,cosinedistance.
Afterexplainingthealgorithmicaspect,youwoulddiscussthedataengineeringside.Proposean
engineeringinfrastructurethatscalestomillionsofusers/shoeswhererecommendationsare
generatedinrealtime.Asanexample,youcanstreamtheuserdatatoaS3bucket.Youcan
performthematrixanalysisonanightlybasis,precomputetheentiresetofrecommendationson
aperuserbasis,andstorethisinainmemorydatabasesuchasRedis.Thenyoucouldbuilda
RESTAPIthatwouldquerythedatabaseandrespondwiththerecommendationsgivenauserid.
Question2.HowmuchisthemonetaryvalueofashareofaChange.orgpetitiononfacebook?
(Change.orginterviewquestion)
www.springboard.com
ULTIMATE GUIDE TO DATA SCIENCE INTERVIEWS
52
Rationale:Theintentofthisquestionistoseehowmuchyouunderstandaboutthebusinessand
howwellyoucanbreakafairlycomplexproblemdowntobasicconceptsandthenconvertthese
conceptstoanalyzablechunksbasedontheavailabledata.Thisisagoodtesttoseehowwellyou
canabsorbacompany’sframeworkforthedataandhowwellyoucancommunicatebusiness
insightsderivedfromyourdataanalysis.
HowtoAnswertheQuestion:Makesureyouresearchthecompaniesyouinterviewfor
thoroughly,especiallytheirrevenuemodel.Getasenseforwhatimportantmetricsthecompany
wouldusetotrackitsperformance,andgetusedtothinkingaboutwhatactionsacompanymust
drivetomakerevenue.Askquestionsandstateanyassumptionsyoumighthave,whichsketchout
howyou’rethinkingaboutthisproblem,thenanswerwithforceandconvictionasifyou’re
presentingtoyoursupervisor.
TheAnswer:ThisquestionrequiressomebasicunderstandingoftheChange.orgbusiness.A
shareofapetitioncanresultinrevenuegenerationintwodifferentways
1. Anotheruserclickingonanadvertisement(i.e.signingapaidpetition)
2. Anewusersigninguponthesystemwhothengoesontoclickonasetofadvertisements
duringthatuser’slifetime
Thefirststepisfiguringoutamethodologythatwouldallowyoutoderiveavalueofbothofthese
ways.Thetrickistostartsimple.Youcansimplifythevalueequationtothefollowing:
Valueofashare=Expectedrevenuefromclickinganad+Averagenumberofnewsignupsper
shareevent*LifetimeValueofanewsignup
Expectedrevenuefromclickinganad=Likelihoodofanadvertisementclick*Averagecostper
clickchargedtopublishers
Likelihoodofanadvertisementclickcanbederivedbyjustlookingatthehistoricaldataand
findingtheaverageconversionrateoverthecourseofatimewindowsuchasamonthoryear.A
similarvaluecanbederivedforthecostperclick.
www.springboard.com
ULTIMATE GUIDE TO DATA SCIENCE INTERVIEWS
53
FortheLTV,it'salittletricky.Youneedtolookatusersoverthecourseofsimilarlifetimesand
derivetheirtotalrevenuegenerated.Onecommonmethodofdoingthisiscalledthecohort
analysisorretentionanalysis.Youcangroupusersthatsigneduponaspecificmonthandlookat
howmanyofthemclickedhowmanyadsoverthecourseofthenexttwelvemonths.Dothisover
twelvedifferentcohortmonths,andthentaketheaveragerevenueoverthelifetime.Now,the
lifetimetoanalyzecanbesettobehoweverlongittakesthatcauseandeffectrelationshiptobe
considerednegligible,i.e.theuserthatsignedupduetotheinitialsharewouldhavesignedup
anywaybeyondthattimewindow,hencetherevenuegeneratedcannotbesolelyattributedtothe
share.
OnceyouhavetheLTV,plugitintotheoriginalequation,andyouhavethevalueofapetition
share.Therearedeeperelementsyoucangointo,suchastherevenuegeneratedbythenewly
joininguserssharingthemselveswhichcausesotheruserstojoin.Makesurethatifyouaregoing
toincludeadditionalelementstoyouranswerthatitdoesn’tdiluteyourmainmessage.Stay
laserfocusedonansweringtheoriginalquestion.Ifyouhaveassortedthoughtsonthesituation,
leavethemtotheend.
3Givenasetofhistoricalnewsarticlesthathavebeenclassifiedasspecificcategoriessuchas
Sports,Politics,World,howwouldyouclassifyanewarticle?
Rationale:Thisquestionlooksathowdeeplyyouunderstandthedatasciencemethodologyand
yourexperiencewithdealingwithunstructuredtextdata,animportanttestforhowcomfortable
youarewithdataformatsthatmightbedifficulttodealwith.
HowtoAnswertheQuestion:Specifyhowyouwouldorganizethetextandhowyouthinkof
classificationsystems.
Samplesolution:
1. Explorethedataandunderstandkeyelementsofthedata.
a. Plotthedistributionofvariouscategoriesinyourtrainingsettodetermineifthereis
labelimbalance.
www.springboard.com
ULTIMATE GUIDE TO DATA SCIENCE INTERVIEWS
54
b. Lookatthetexttoidentifyanythingstrange,suchasnonenglishtext,heavy
abbreviations,ormisspellings.
c. Dotopicextractiontoidentifykeywordsforspecificlatenttopicsandfindcorrelation
tothelabelledcategories.Thismaygiveyouahintastowhethertherearelatent
topics(keywords)thatmaycorrelatebetterthanjustusingallthewords.
2. Derivethetrainingsetbycleaningupthetext.Removelesserinformativeelementssuchas
punctuation,abbreviations,andunicodecharacters.Dofurthercleaningbytakingthelower
caseofwordsandlemmatization/stemming.
3. UseaTFIDFvectorizertoconvertthedatatoabagofwordsmodelwithTFIDFmetric.et
lowerandupperboundstoTFIDFtoreducethevocabularysize.
4. Buildapipelinewhereyoucantrainvariousmodelsandcomparetheirperformance
relativetometricssuchasAUC,F1score,precision,andrecall.Youcandogridsearchto
automatethecrossvalidationaspectaswell.
5. Onceyougettheoptimalmodel,youcanpublishthismodeltoproductionusingapickled
model(inpython)orPOJO(injava).Thismodelcanthenbequeriedbyusingtheexact
sameprocessofcleaningasdonein#2and#3forthenewarticles.
4Designanexperimenttofigureoutwhichwebdesignalternativetouse.Assumetherehave
beennootherexperimentsdoneandthereisnoknowledgeoftheuserbehavior.Discusspotential
issuesthatcanoccurwiththeconclusionsandhowtoavoidthem.
Rationale:Manywebcompaniesaskthisquestionbecauseitistheirbreadandbutterto
optimizetheirwebsiteforbetterbusinessresults.ThinkofFacebookconstantlychangingtheir
homepagetogetyoutopostmore.Thedatascientist’sroleisofteninhelpingtheproductmanager
setuptheexperimentorinterprettheexperimentresults.Thegoalofthequestionistoseethe
depthoftheknowledgeoftheintervieweeinthistopic.
Solution:Identifythenatureofthechangeandthemetrictoconsidertodecidewhichversionof
thesitetochoose.Forexample,clickthroughrateandaveragenumberofFacebookshares.
Next,decidethenumberofsamples/visitsnecessarytohitthenecessarystatisticalsignificance
(e.g.95%).Thiscanbedonebyusingachisquaredtest(ifweareusingabinomialrandom
www.springboard.com
ULTIMATE GUIDE TO DATA SCIENCE INTERVIEWS
55
variableofclickingvs.notclicking)oraztest(ifweareusinganormallydistributedrandom
variable).YoucanthenevaluatethepvaluetoidentifywhetherthemetricoftheBtestis
statisticallysignificantlydifferentthanthemetricofthebaselineAtest.Ifitisandthemetricis
betterthanthebaseline,thenthealternativesiteisthebetterwaytogo.
Someotherissuesyoushouldconsiderinthisanswer:
1) Identifypotentialbiasesduetointeractionsacrosspages.Talktotheproductmanagerand
seeiftherearewaysthatarandomsamplingmaynotworktotestthenatureofthechange
you’reproposingforawebpage.
2) PerformaA/Atestwhichimpliestestingtworandomsamplesofvisitors,andcheckifthe
distributionandmetricofchoicedoesnothaveastatisticallysignificantdifference.This
willensurethefairnessoftheA/Btest.AnA/Atestensuresthatyouraudiencedoesn’thave
aparticularskeworbiasandarandomizedselectionforanA/Btestwillbestatistically
relevant.
3) Whatifthemetricthatweareevaluatinghassignificantoutliersthatmaycausetheaverage
tobeapoormetric?Thedistributionmaybehighlyskewed.Weassumetheaverageisa
goodmetricofcomparisonsincecentrallimittheoremholds.Thismaynotbetrue.Hence,
checkthedistributionofthemetrictoensurethattakinganaverage(e.g.conversionrateor
averagenumberofsharesperuser)isareasonablemetricwhencomparingbetween
alternatives.Ifoneuserhasthousandsofsharesattributedtotheiraccount,forexample,
usingsharerateperusermaynotbethebestperformancemetric.
Insummary,casequestionsaredesignedtotestforyourexperienceandyourknowledgein
differentfieldsofdatascience.Theyaredesignedtoseeifyouhaveanylimitstoyourability.
Demonstrateyourknowledgethoroughly,andyou’llcomeoffwellinanycaseanalysis.
TacklingtheInterview
1) Dressedaccordingly.Ifit’saninterviewforastartup,adressshirtwillsuffice.Ifit’san
interviewwithabank,wearsuitandtie.Ifyou’reunsureofwhattowear,ask.
www.springboard.com
ULTIMATE GUIDE TO DATA SCIENCE INTERVIEWS
56
2) Beforeyoucomeintotheinterview,researchyourinterviewerandthecompany.Comeup
withgoodquestionstoask.
3) Beatthetopofyourgamementally.Eatwell,behydrated,exercisewell,anddowhatever
youcantomakesureyou’repreparedtohandleaninterview.
4) Answerquestionsindetail,andsketchoutyourthoughtprocess.
5) Smile,andbeconfident.Don’tcomeinstressed.Meditate,stretch,orreaddowhateverit
takestogetyoutoyourpeak.
Conclusion
Thedatascienceinterviewprocessisamultifacetedbeast.You’llbechallengedtoprogramand
comeupwithtechnicalalgorithmsonthespot.You’llbechallengedaboutyourstatisticaland
mathematicalknowledge.You’llbechallengedonyourabilitytoleadteams,communicate,
persuade,andinfluence.
Itcanbehardtoseehowtopassthisbeastofaninterviewprocess.Thankfully,wecondensed
actionableinsightsfromsuccessfulapplicantsandthehiringmanagersontheothersideofthe
table.
www.springboard.com
ULTIMATE GUIDE TO DATA SCIENCE INTERVIEWS
57
What Hiring Managers Are Looking
For
InterviewwithWillKurt(QuickSprout)
Bio:WillKurtisaDataScientistwithQuickSprout.His
maininterestsareprobability,writing,andHaskell.He
blogsatCountBayesie.comandcanbefoundonTwitteras
@willkurt
Whatdoyoulookforwhenyou’rehiring
candidates?
Thebiggestthingformehasalwaysbeenacombinationof
creativityandgenuinecuriosity.Inastartupenvironment,
newproblemscomeupeverydayinawiderangeofareas.
Onemonthyoumaybehelpingtheproductteamaddnewfeatures.Thenextmonth,you’llhelp
salesimprovetheirprocess,andthemonthafter,you’llbehelpingmarketingrestructuretheir
testingsetup.Themostvaluablecandidatesaretheonesinterestedinallofthecompany’sdata
relatedproblemsandalwaysthinkingofnewandinterestingwaystosolvethem.
What’sthebestpieceofadviceyoucangivetopeoplegoingthroughthedatascience
interviewprocess?
Inmyexperience,allsmallcompaniesandstartupsworthworkingforareexcitedabouttheidea
ofaddinganewdatascientisttotheteam.Theyhopeyourskillsandexperiencewillhelpthem
solvearangeofproblemsthey’vebeenstrugglingwith.Showuptotheinterviewreadytolistento
whatthey’retryingtosolveandgetthemexcitedaboutsolvingproblemstogether.Everychance
yougetaskpeoplewhatthey’reworkingonandgetthembrainstormingwithyouaboutwaysyou
couldmaketheirdaybetter.Therearethousandsofcandidatesouttherewithsuperbquantitative
www.springboard.com
ULTIMATE GUIDE TO DATA SCIENCE INTERVIEWS
58
skills,butcandidateswhocareandareexcitedareveryrare.Leavetheinterviewwitheveryone
wantingtoworkwithyouonaproject,andthey’llbetheoneshopingyousay“yes.”
Whatkindofinterviewquestionsdoyouliketoask?Whatareyoutryingtotest?
AllIcareaboutishowyourmindworksonceit’sfixeditselfonaninterestingproblem.At
Kissmetrics,Igaveoutanopenended“homework”assignment.Therewasanobviousapproach
totheproblem(buildaclassifier),butImentionedthisandcautionedthatpartofthetestwasto
seeifyoucouldcomeupwithsomethinginteresting.Theresultsoftheassignmentdidn’thaveto
belongorcomplicated.Whatmatteredisthattheystartedaconversationandshowedthatthe
candidatehadgenuinecuriosityinfindingsomethingworthtalkingabout.Giventhatacandidate
cancodeandiscomfortablewithlinearalgebra,calculus,andprobability,theyhavethebasicsto
learneverythingelse.Itisveryhardtoteachsomeonetothinkcreativelyorbecomepassionate
aboutproblems.
WhatisdifferentabouthowKissmetricsandQuickSprouthiredatascientists?
Rightnow,Quicksproutisaverysmallteamintheearlystagesofproductdevelopment,sowe’re
nothiringnewdatascientistsatthemoment.Onethingthataspiringdatascientistsshouldknow
isthatmanystartupsandsmallcompaniesarelookingforadatascientistbutmayhavegivenup
onfindingoneasthesearchprocesscanbeexhausting.OneofourbestcandidatesatKissmetrics
showedupatourdoorandsaid,“Iwanttoworkhere!”Peoplecomingfromacademiaorother
largeorganizationsmightnotbeawareofhowflexiblestartupsandsmallcompaniescanbewhen
itcomestohiring.Ifyouthinkacompanyisdoingcoolwork,connectwiththem.It’shardtomake
abetterimpressiononagroupofpeopleexcitedabouttheirworkthantellingthemyoulovewhat
they’redoingandwanttobeapartofit.Evenifthatcompanyisn’thiring,you’llbeatthetopof
thelistif/whentheydostartlooking.
 
www.springboard.com
ULTIMATE GUIDE TO DATA SCIENCE INTERVIEWS
59
InterviewwithMattFornito(OpsVisionSolutions)
Bio:MattFornitoisaDataScientistandLeaderwithover
tenyearsofexperienceintheresearch,analytics,and
managementdomains.Apassionforlearninganddevout
workethiccontinuestohelphimgrow.Thisinterviewis
transcribedfromnotestakenonaphonecallwithMatt.
Whatdoesyoulookforwhenyou’rehiring
candidates?
Ifeelmostcomfortablehiringpeoplewithastrong
quantitativebackgroundwhocanlearnprogramming
ratherthantheotherwayaround.AMastersoraPH.Disveryimportanttome,asIfeelthat
undergradisnotastrongsignalofsuccess;it’sarelativebreezeformostpeople.Ipreferhiring
peopleabletopickupprogrammingandeffectivecommunicationknowingandunderstanding
whatthetechnicalproblemsaretoimplementingasolutionandbeingabletocommunicatethose
conceptsiskey.Whatdifferentiatesdatascientistsanddataanalystsistheabilityofdatascientists
todeeplyunderstanddataproblemsandhowtosolveforthem.
IlikerecruitingMastersandPhDsfrommathandstatistics,chemistry,physics,and
bioinformaticsandengineering.ThereareasmallhandfulofpeopleinMBAsthathaveworked
outgreatforme.IamactuallyaPhDinorganizationalpsychology,sothoughItendtotrytohire
peoplewithSTEMbackgrounds,itisn’tastrictlimitation.
What’sthebestpieceofadviceyoucangivetopeoplegoingthroughthedatascience
interviewprocess?
RecruiterslookateducationlevelandthelasttwojobsontheCVandtheirpedigree.HRsonly
takeaveryquickglanceatCVs,soyouhavetostandoutinamatterofseconds.Onepieceof
advice:getyourselfintoabigcompanythathasapedigreelikeFacebook,orgointoastartupand
takeahighpositionsothatyoucanstandouteasilyforadvanceddatascienceroles.
www.springboard.com
ULTIMATE GUIDE TO DATA SCIENCE INTERVIEWS
60
“Walkmethroughaproject”questionswhereahiringmanagerwillaskexactlyhowyoubuilt
somethinginthepastarehugeeverythingfromwhatdatawasused,whattoolswereused,what
theoutcomeswereareimportanttorecountclearly.successfulintervieweeshaveacomfortable
grasponwhatthey’veworkedonandarereadytostorytellonthatelementandrelatehowtheir
workimpactedthebusinesstheywereworkingfor.
Whatareyoutestingfor?
QuestionsIaskinvolveworkingaroundaprojecttotestproblemsolvingandcommunication
skillsacrosstheinterview.Iamalsoassessingacandidate’spassionforthecompanyanddata
science.Adriveforcontinuouslearningandloveofproblemsolvingarekeydifferentiators.Then
onthetechnicalside,Iaminterestedinseeingcandidatesworkonhowtooptimizedatawith
HadoopandSparkandworkingonthetradeoffsbetweendifferentdatasciencesolutions.Dothey
thinklikeadatascientist?Havetheydonedatasciencework?TheseareimportantquestionsIam
lookingtouncoverwithmyinterviewprocess.
Iwillthengointomathquestionssuchasaskinghowgradientdescent,statisticaltechniques,and
randomforestwork.Acoupleofsituationalquestionswherethecandidateisputthrougha
hypotheticalclientsituationaredeployedtoseehowthecandidatewouldhandleinterfacingwith
clients.IhaveastrictrequirementofabilitytoprograminPythonorR,butIamflexiblewithC++
andJava.Idon’tbelieveinHackerRankliketestingsituationswhereyouareexpectedtotraceout
asolution;Iwouldrathertestforadaptiontonewprogramminglanguagesandanabilitytolearn
skillsrapidly.Anybodyhiredisgoingtohavetohavethelatentskillofadaptability,andthatis
thekeythingIamtestingfor.
 
www.springboard.com
ULTIMATE GUIDE TO DATA SCIENCE INTERVIEWS
61
InterviewwithAndrewMaguire(PMC/Google/Accenture)
Bio:AndrewhasbeenworkinginAnalytics/Data
Sciencefor7yearsinvariousrolesacrossmany
differentindustries.HeisaDataScientistatPenske
MediaCorporationfocusingonbothdata
engineeringinfrastructureaswellasapplied
businessanalytics.Priortothisposition,heworked
atGoogle(marketinganalytics,thenlocaldata
quality),Accenture'sAnalyticsInnovationCentre
(consultancy),andAon'sCenterforInnovationand
Analytics(productdevelopmentteam).
Whatdoyoulookforwhenyou’rehiring
candidates?
Beyondmeetingthebasicrequirementsfromatechnicalandexperiencepointofview,I'dsay
enthusiasm,willingnessandabilitytocontinuallylearnnewthingsarekey..
Agoodattitudeissuperimportant,sosomeonewhoisabletoalsotellmeabouttheirweaknesses
aswellasstrengthsisagoodwaytodrawthisout(sometimessellingtoohardisabitoffputting;
humilityismuchbetter).
Beingapproachable,openandhonestissomethingthat'skeyonthe'teamfit'side.Youdon’thave
toknowtheanswerforeverythingbutbeingabletoworkwithotherstocomeupwithadecent
solutioniscrucial. 
What’sthebestpieceofadviceyoucangivetopeoplegoingthroughthedatascience
interviewprocess?
Onthetechnicalstuff,takeyourtime,writestuffdown,andaskclarifyingquestions.Alsodon'tbe
afraidtotellthemifit’sanareayou'venotworkedonbeforeoranalgorithmyou'renotthat
www.springboard.com
ULTIMATE GUIDE TO DATA SCIENCE INTERVIEWS
62
familiarwith.Beingabletoadmitwhenyourknowledgeislimitedissuperimportantasadata
scientist;continuallylearningisoneofthemostimportantskillsrequired.
Makesureyouhavetwoorthreedatascience'stories'youcanchataboutwithaninterviewerthat
touchonproblemformulation,datawrangling,analysisandinsights,visualizationand
stakeholdercommunication.Trytogetthebalancerightbetweencoolnerdytechnicalstuffand
showingbusinessunderstandingandinsights.These'stories'canbeprojectsfromyourprevious
roles,collegeassignments,orprojectsyoudidonyourowntime.Getgoodatspottingopenings
frominterviewerquestionstouseyourstoriestoshowconcreteexamplesandexperience.Ifind
thatchatting(indetail)aboutprojectsthecandidatehasdoneinthepastisthebestwaytogeta
properfeelforthem(andbestplacetoprobedeeperfrom),somakesureyoumakeiteasyforthe
interviewertobeinterestedandexcitedtoaskyouaboutsomeproject’sorexample’sfromyour
CV.
Whatkindofinterviewquestionsdoyouliketoask?Whatareyoutryingtotest?
What'sthebiggestormostcomplexdatasetyouhaveeverworkedwith?Whatproblemsdidit
create?(Tryingtobeginadiscussionherethatcanleadintojudgingdatawranglingskillsand
experience)
Givemeanexampleofatimewhenyouanalysedadataset,andcommunicateyourfindingsback
tothebusiness.Whatwastheproblemfaced?Whatdidyoufind?Howdidthisaffectthe
business?(Touchontheextractingbusinessinsightsandcommunicatingbacktostakeholders
aspects)
IaskquestionsveryrelatedtowhatisontheCV,soifit'saprojectfromapreviousrolefor
example,Iwantyoutoexplainwhattheproblemwas,whatsortofdatayouused,howyouusedit,
whattheinsightswere,andhowthisallfitsintothewiderbusiness.Choosewhatyouputonyour
CVverydeliberately.Ifyoufindithardtogetallontwopagesthenmaybehavedifferent‘types’
ofcv’syoumightusefordifferenttypesofroles.
Finally,Iaskcandidatestogivemeanexampleofatimewhentheyfailed,thenaddwhatthey
thinkwentwrongandwhattheywoulddodifferentlyinfuture.Thisissomethingthatcomesout
ofHR101,butIliketohearwhattheyhavetosay:)
www.springboard.com
ULTIMATE GUIDE TO DATA SCIENCE INTERVIEWS
63
WhatisdifferentabouthowGooglehiresdatascientistsfromtherestofthe
industry?
I'mnotsurethereistoomuchofadifferenceanymore.Generallyitdependsonthespecificrole.
Forveryspecializedpositionsthatareoftenmorelikeresearchorfellowshippositions,youwould
getmuchmoredetailedtechnicalquestionsandproblemstodriveintotherelevantareaof
expertiseinveryfinedetail.Formoregeneralistorbusinessrelatedroles,thefocusismoreonthe
rightmixoftechnicalskills,businessunderstanding,workinginteams,andcommunicating
resultstostakeholders.
ThemaindifferenceinGoogleisthatyouhavealotmoreinterviewsandmeetmorepeople,so
behindthescenestherearearound6+peoplewhohaveallmetyouandprobedyoufromtheir
owndifferentangles.Thesepeopleallhaveadifferentviewofyouandyourstrengthsand
weaknessesandmustcometoadecisionandconclusiontogetherthattypicallyinvolvestradeoffs.
Beingabletoshowdecentlevelofcompetenceacrosstheboardasopposedtobeingarockstarin
oneareabutlettingyourselfdowninotherswillgenerallyserveyouwell.Thisiswhereattitude
andbeingeasytogetalongwithcanbemostimportant;evenifyoufallalittleshortononeofthe
competencies,iftheylikeyouandfeelyoucouldeasilygetuptospeedinthatareainafew
months,thenit'slesslikelytobeadealbreaker.
 
www.springboard.com
ULTIMATE GUIDE TO DATA SCIENCE INTERVIEWS
64
InterviewwithHristoGyoshev(MasterClass)
Bio:HristoGyoshevistheHeadofBusinessOperations&
StrategyatMasterClass,afastgrowingstartupthatis
democratizingaccesstogeniusandreimaginingonline
education.Hepreviouslyworkedoncorporatestrategy,
businessoperations,andproductstrategyatboth
consumerweb(e.g.Yahoo!)andenterpriseSaaS
companies.MasterClassislookingtohireaDataScientist
&manyotherpositions.Checkoutthedetailsat
careers.masterclass.com/
Whatdoyoulookforwhenyou’rehiring
candidates?
Oneofthemainassetswelookforisadesiretoworkonprojectsacrossaverybroadrangeof
analyticdisciplines–fromquantitativemarketresearchand/ordesigning,conducting,and
analyzingusersurveys,tostatisticalanalysis,tobusinessintelligenceandanalytics.Wealsolook
forcandidateswhoarecomfortablelearningsomethingnewtoremovebottlenecksandkeepa
projectmoving,whennecessary.
Intermsofeducationalbackgroundandexperience,we’relookingforananalyticalbackground
thatcombines1.sufficientknowledgeofstatisticstodeterminewhatisorisn’tavalidstatistical
inference,recognize&preventbiases,etc.;and2.thedesireandabilitytoobtainandworkwith
realworlddata(whichisalwaysimperfect)andderiveactionableinsights.
Someonewhohasaverystrongquantitativebackgroundandabilitytoprocessandanalyzedata
usingExcel,SQL,andPythonorR;whoalsohasexperienceinsocialscienceresearchor
market/userresearch(througheitheracademicorindustrywork);andwhohasexperiencewith
businessreporting/analytics,couldbeanidealcandidateforus.
What’sthebestpieceofadviceyoucangivetopeoplegoingthroughthedatascience
interviewprocess?
www.springboard.com
ULTIMATE GUIDE TO DATA SCIENCE INTERVIEWS
65
Strivetounderstandandkeepinmindthebroadercontextoftheproblemyouarebeingaskedto
solveortheproblembehindthequestionyouarebeingasked.Wheneveryouareaskedto
performacertainanalysis,orbuildamodel,someoneatthecompanybelievesthatthiswouldhelp
themsolveaparticularproblem.Sometimesyoucantellinadvancethatitwon’t,andsometimes
youcansuggestabetterapproach.Youranalysis/model/otherworkproductwillalwaysbebetter
ifyoustartfromagoodunderstandingofthecoreobjectivesofthe‘clients’ofyouranalysis.(This
appliesasmuchtoquestionsyouareaskedduringtheinterviewasitdoestoprojectsyouare
askedtoworkononceyougetthejob.)
Whatkindofinterviewquestionsdoyouliketoask?Whatareyoutryingtotest?
Weliketounderstandacandidate’spreviousexperiencewithvarioustypesofworkthatweexpect
willberelevanttotheirrole.Thus,wemayaskforexamplesofspecifictypesofprojectstheyhave
workedon,andthenaskthemtowalkusthroughtheirapproachandthinking,thetoolstheyused,
themajorchallengestheyencountered,andhowtheyresolvedthem.
Wemayalsoaskcandidatestocompleteashortprojecttoseehowtheyapproachsomespecific
problem–andyes,tobeabletoseethequalityofadeliverabletheyproduce.
WhatisdifferentabouthowMasterClasshiresdatascientists?
ComparedtomostDataScienceroles,thejobwithusinvolvesverylittlemachinelearningor
algorithms,andonlyminimaldatawrangling,butaverywidevarietyofanalysesthatwould
informabroadrangeofdecisionsabouttheproducts,business,andoperationsofthecompany.
Theworkwould,ofcourse,involvesomeexporting,processing,andanalyzingdatafromvarious
systems,butwouldalsoinvolvebuildingvariouspredictivemodels;designing,conducting,and
analyzingsurveysorexperiments;helpingtodefineandsetupreporting&metrics;and
conductingoneoffanalysesrelatedtovariousaspectsofourbusinessoperations.
Correspondingly,wedon’tneedcandidatestobeproficientinmachinelearningoralgorithms,but
wedoneedthemtobehighlyversatileandfamiliarwithanumberofotheraspectsofdata
analysis.Wealsoneedthemtobewillingandabletolearntoolsormethodstheymaynothave
previouslyused.
www.springboard.com
ULTIMATE GUIDE TO DATA SCIENCE INTERVIEWS
66
Conclusion
Hiringmanagersacrosstheboardloveitwhenyoudemonstrate:
1) Passionforthecompanyanddatascienceingeneral
2) Anabilitytogetalongwellwitheverybody,whichmayevenhelpyouwithweaknessesin
yourtechnicalability
3) Strongwillingnesstolearnanddemonstratedabilitytorapidlydoso
4) Astrongrecordofpreviousprojectsandtheabilitytorelatepreviousprojectswithimpact
driven
5) Stronganalyticalability
Nowlet’stalkabouttheothersideofthetable:successfulapplicantswhonowworkasdata
scientists.
 
www.springboard.com
ULTIMATE GUIDE TO DATA SCIENCE INTERVIEWS
67
How Successful Interviewees Made It
SaraWeinstein
DataScientistatBoeingCanadaAeroInfo,
Springboardgraduate
What is advice you’d have for how to ace
thedatascienceinterviewprocess?
In terms of preparation, I wish I spent more time
thinking about analytics strategy. I prepped hard
on stats, probability, ML, python/R...all the
technical stuff, but was nearly caught off guard
by a straightforward question about how I'd
approach a particular problem given a specified
data set. My answer wasn't as confident as I
would have liked. I'd been so focused on the
"hard" stuff that I hadn't thought that much
about higherlevel analytics methods &
strategies.
What surprised me and what I found
difficult:
How long the process took. I knew to
expect several interviews, and in fact had
three. With nearly a week between each,
plus waiting for my background check to
clear, the process from first contact to firm
offer took a month. It was stressful to say
the least. Staying positive, confident, and
prepared for a whole month was
challenging. It would have been much
easier to bear if I'd known in advance that
it would take that long. For others facing a
lengthy multiinterview hiring process:
meditation is your friend. It helped me
sleep at night, and I used the techniques
right before interviews to channel calm and
confidence.◼
www.springboard.com
ULTIMATE GUIDE TO DATA SCIENCE INTERVIEWS
68
NirajSheth
DataAnalystatReddit,Springboardgraduate
What is advice you’d have for how to ace
thedatascienceinterviewprocess?
I wish I had studied more fundamental statistics
before interviewing. It's silly, but people often
look for whether you are familiar with terms like
Type I and Type II errors. Depending on the time
you have, I suggest getting a statistics textbook
and at least becoming familiar with the terms out
there.
I should have probably expected this, but I was
surprised how poor we are as an industry in
evaluating projects. When I talked about past
projects, everyone just cared about interest value
(does the analysis say something interesting?) 
nobodyquestioneddeeplythemethodsIused.
You didn't ask this, but there were also
some things I did that I think worked out
well. One is to have a live project up
somewhere with a neat visualization (i.e.
more than a github repo with a readme). It
doesn't have to be fancyjust prove you can
build something that works (mine was a fog   
prediction map, for example). It definitely  
helpsgetyourfootinthedoor.
The other thing is to ask for a takehome
data set. I don't know about you, but I've
found that for myself and other people who
don't have a formal data background, it can
be intimidating to work on a data set on the
spot; I just hadn't developed the muscle
memory for it yet. However, I knew the
right questions to ask, and I could figure
out how to answer them if I had a little
time, so getting a takehome set let me
showwhatIcoulddothatway.◼
www.springboard.com
ULTIMATE GUIDE TO DATA SCIENCE INTERVIEWS
69
SdrjanSantic
DataScientistatFeedzai|DataScienceMentor
atSpringboard
What is some advice you’d have for how
people can ace the data science interview
process? What were some of the toughest
questions?
The most important thing, in my opinion, is
understanding how the major supervised and
unsupervised algorithms work and being able to
explain them in an intuitive way. A good
command of Data Science terminology is crucial.
Candidates should also have a thorough
knowledge of relevant accuracy metrics, as well
as the various approaches to evaluation
(train/test, ROC curves, crossvalidation). The
tougherquestionswould
relate to these same affairs, but with having
tobreakoutthemathonawhiteboard.
Howdidyourinterviewprocessgo?
Luckily, very smoothly! Most of my
interviews had a feeling of being a
conversation between peers, so I didn't find
them very stressful. The companies I
interviewed with moved very quickly (one
round a week), which helped streamlined
the process. I was also very impressed as to
how most companies that turned me down
gavemeveryhonestfeedbackastowhy!
What were some of the factors for
youinchoosingyourcurrentjob?
Primarily, it was the opportunity to use a
technical toolset and solve problems I
hadn't solved before. My previous role was
very focused on just building models. The
data was already completely cleaned and
preprocessed, and the exploratory work
was done using a commercial GUIbased
tool. I felt that my datawrangling and
command line edge was being dulled slowly
and jumped at the opportunity to work in
an environment where I'll be able to "get
myhandsdirty"oncemore!◼
www.springboard.com
ULTIMATE GUIDE TO DATA SCIENCE INTERVIEWS
70
Conclusion
Thecommonpointsforsuccessthesedatascientistsbringtotheforefrontareasfollows:
1) Don’tthinkquestionsaboutbasicmaterialwon’tbecovered.Readuponstatistical
fundamentalsbeforeyougothroughtheinterviewprocess.
2) Bepreparedtodowellonnontechnicaldimensions.Companiesaretestingyouonyour
communicationskillsandyourabilitytogetalongwithfuturecoworkersasmuchasthey
aretestingyouonyourstatisticalandprogrammingknowledge.
3) Bepreparedtostorytellaboutwhoyouareandwhyyourpassionsandskillsareuniquely
valuableforthecompanyathand.Havingrelevantprojectsandbeingveryclearaboutwhat
youcontributedtothoseworkswillmarkyouasacandidateworthyofpassingtothenext
round.
4) Bepatient.Aninterviewprocesscantakealongtime.You’llwanttobepreparedtowait.
We’veprovidedyouallthatwehaveontheactualdatascienceinterviewprocess.Nowwehaveto
lookatwhathappensafteryou’vefinishedinterviewing.
 
www.springboard.com
ULTIMATE GUIDE TO DATA SCIENCE INTERVIEWS
71
7 Things to Do After The Interview
Afteryou’vefinishedyourdatascienceinterview,youmightthinkyourworkisfinished.That’snot
necessarilythecase.Herearealistofthingsyoucandoaftertheinterviewtoensure,asbestas
possible,thatyoumaximizeyourchancesofmakingthebestlastingimpressiononyourpotential
employers.
1- Send a follow-up thank you note
Itisnowcustomarytosendafollowupthankyounote.Mostrecruitersnowagreethatitis
mandatorytodoso.Witheachofficeworkerreceivinganaverageof110emailsaday,youwon’t
wanttojuststickwithaboilerplate“Thankyoufortheopportunity”email.Howyoufollowupon
aninterviewcanmakethedifferencebetweeninternaladvocatesfightingtogetyouin,and
apathy.
Makesureyou’reremembered.You’llwanttosendanemailattheveryleast.Candidateswhotake
theextrastepofsendinghandwrittennotesoralistofthoughtsaftertheinterviewwillstandout
fromtherestoftheaverage109emails.
2- Send them thoughts on something they brought up in the interview
Oneeasywaytodifferentiateyourselfistogobeyondsayingthanks.Rememberwhathas
happenedintheinterviewandmakeaconsciousefforttoteaseoutexactlywhatpainpointsthe
employeristryingtosolve.Ifsampleproblemswithintheinterviewareorientedtowardsa
technicaldirection,oraquestionnotesadisconnectbetweendifferentteams,you’llwanttomake
anoteofitandsendindepththoughtsonanycompanyproblemsthatmayhavesurfacedduring
yourdiscussion.
www.springboard.com
ULTIMATE GUIDE TO DATA SCIENCE INTERVIEWS
72
Afterall,aninterviewisn’tjustatest;it’sadiscussion.Ifyoulistencarefullytothequestions
presentedandasktherightquestionsyourself,youwillknowexactlywhatproblemsthecompany
isfacing.Makesureyousendthemthoughtsonwhatsolutionsyou’dpursue.
3- Send relevant work/homework to the employer
Itcanbedifficultseeinghowyourdifferentskillsapplytotheoffice,especiallyforsomebodywho
hasjustmetyou.Thesharpesthiringorganizationswilloftengiveyouasampleproblemtosolve
thatissourcedfromsomerealissuetheyarefacingrightnow.Thisgivesyouthechanceto
demonstratehowyoureffortscanimpactthebusinessinapositivemanner.
Organizationsthatdon’tdothatwillhesitatetohiretherightcandidatebecausetheyhaven’t
sufficientlydemonstratedhowthey’ddriveimpactforthecompanyinquestion.However,youcan
beproactiveandusewhatyoulearnedintheinterviewtofollowup.Youdon’thavetostopat
sendingthemthoughtsthatshowyoulistenedcarefully;youcangivethemactual,tangible
solutions
TheauthorofthispostonForbeswastoldthattheydidn’thaveenoughofaportfoliotogetajob
asafreelancecopywriter.Aftertheinterview,thehiringmanagertoldthemthattheylikedthe
spiritthecandidatehad,butwerehesitantduetoalackofaportfolio.Havinglistenedcarefully
throughouttheinterview,thecandidateknewthatamajorproject(theredesignofawebsite)was
justoverthehorizon.
Insteadofacceptingdefeat,thecandidatesenttenproposedheadlinesforthewebsitebanner,free
ofcharge.Thisburstofinitiativegotherthejobofdoingtherestofthewritingforthe
websiteandtheattentionofaverybusyemployer.
Youneedtohaveaportfoliothatshowstheimpactyoucanmake,butsometimesthatisn’t
enough.Ifyou’reastuteandyouasktherightquestions,youcanfindamajordataproblemforthe
company.Therealwaysissomethingthat’swhythey’rehiringforthefirstplace!There’sadata
projectouttherethateverybodywouldlovetoseedoneorathornyproblemthatnoonecanfigure
out.
www.springboard.com
ULTIMATE GUIDE TO DATA SCIENCE INTERVIEWS
73
Sendthemaplanforwhatyou’ddoorplaywithsomeofthedatathey’vedivulged,andgivesome
solidinsightsintohowyouwork.Proactiveinitiativewillgoalongwaytogettingyouanoffer.
4- Keep in touch, the right way
Oneofthemostawkwardpartsofthepostinterviewprocessiswaitingforaresponse.Youdon’t
wanttocomeoffasdesperatebyfollowinguptoomanytimes,butcompaniestaketheirtimeif
youdon’tengagewiththemproactively.
Itispossibletoeffectthepostinterviewdecisionfromoutsideofthecompany,butyoushould
keepinmindtheappropriatechanneltoreachsomebody.Makesuretoaskbeforetheinterview
endshowbesttoreachyourinterviewer.Everybodyhasapreferredmodeofcommunication;if
theyspecifyshortemailsortocheckinonceinawhileinperson,followthatruleanddispelsome
ofthepostinterviewawkwardness.
5- Leverage connections
Youshouldhavecomeinwithstrongreferencesbothfromexternalandinternalsources.Ifyou
hadbeenbuildingyournetworkandprovidingvaluetothem,youshouldhavestrongadvocates
thatcansupportyourcandidacy.Checkinwithpeoplewhohavereferredyouinternallyeveryonce
inawhile,andifneeded,getthemtoadvancehowexcitedyouwouldbetoworkatthecompany
andhowluckythecompanywouldbetohireyou.
Hiringisoftennetworkdriven,andthestrongestsignalyoucansendtoapotentialemployerisa
strongnetworkofpeoplewhoarewillingtogotobatforyou.
6- Accept any rejection with professionalism
Nomatterwhat,you’reoftengoingtogetrejected.Sometimes,you’renotrightfortherole,orthey
mighthavefoundsomebodywhoisabetterfit.It’simportantatthispointtomaintainyour
composure,thanktheemployerfortheirtime,andmoveon.
www.springboard.com
ULTIMATE GUIDE TO DATA SCIENCE INTERVIEWS
74
Peopleintheindustrytalkamongsteachother,andbeingunprofessionalatthispointwillonlybe
badkarmaandmightgetyouignoredatothercompanies.Beingprofessionalensuresthehealthof
yournetwork.Moreimportantly,anoisn’talwaysano.Sometimes,companiesdokeepyour
profileonfileandtheywillreachoutforajobthatistheperfectfitforyou.
PerhapsWinstonChurchillputitbestwhenhesaid“Successistheabilitytogofromonefailureto
anotherwithnolossofenthusiasm.”
J.KRowling,theauthorofthepopularHarryPotterseries,sharedherrejectionlettersfrom
publishers.BrianChesky,thefounderofAirBnB(nowvaluedatmorethan10billiondollars)
publishedsevenrejectionlettersfrompotentialinvestors.Inordertoachievegreatness,youwill
havetoendurerejection.Everybodysuccessfulalreadyhas.
7- Keep up hope
Theinterviewprocesscanbeoneofgreatanxiety.Yourfuturecanbemappedoutbydeciding
whatcompanyyoucanworkfor.Aninterviewcanmeanthebeginningofacareerchange.Itcan
meanmovingcities.Itisaperiodinourliveswhereotherpeoplehaveadisproportionatecontrol
overourdestinies.
Nevertheless,asseenintheprevioussteps,youcontrolalotmorethanyouthink.It’simportantto
keepyourheadupanddowhatyoucan.Themostimportantthingyoucandoduringthe
interviewprocessistokeepuphope.Interviewsarelengthy.Companiestaketimetogetbackto
you.Therearelengthyinternalchecksandprocessesbeforeacandidategetsaccepted.Youmaygo
throughmultipleroundsofinterviewswiththesamecompanyandnotseemanyclosertoafinal
offer.
Youhavetosetexpectations.DJPatil,theChiefDataScientistoftheUnitedStates(aposition
createdforhimbyPresidentObama)tooksixmonthstotransitionoutofacademiatoajobinthe
industry.Youshouldneverbedisheartenedduringyourownjourney. 
www.springboard.com
ULTIMATE GUIDE TO DATA SCIENCE INTERVIEWS
75
The Offer Process
Yourgoalistogetasmanyinterestingoffersaspossiblethatyoucanevaluateandnegotiate.
Whiletheprocessitselfisdifficult,andmaytakelongerthanyoucouldexpect,onceyoustart
gettingoffers,you’llhaveearnedthem.
It’skeytoemphasizehowimportantitistomanageyourexpectationsandkeepyourhopeup.
Severalofthedatascientistsweinterviewedtalkedaboutmonthstohalfayearofwaitingto
transferfromanadvanceddegreefromaprestigiousschooltoasecurejob.Alotofthemhadto
takeentrylevelpositionstogettheirfootinthedoor.
Youmighthaveheardalotofgreatthingsaboutdatascience,butyou’llonlyexperiencethatwith
alotofhardworkandwaiting.
Makesureyouweighwhatispresentedtoyouandchoosethefutureyoudeserveonceyou’ve
spentallthehardworkearningit. 
HandlingOffers
Ifyoufinishaprocesssuccessfully,youmighthaveoneofferormultipleoffers.Congratulations!
Acceptinganofferisacommitmentofsignificantamountsofyourtimetothecompanyin
question.Alwayskeepthatinconsideration.Thereareseveralfactorsyoucanusetoascertain
whetherornotanofferistherightoneforyou.
Company Culture
Thismightbeoneofthemostimportantfactorsindeterminingwhenanofferisoneyoushould
accept.Makesureyouaskaboutthekindofcompanycultureyou’regoingtobeapartof.Lookfor
signsthatthecompanyhasindividualsthatgenuinelyenjoyspendingtimewithoneanother,and
www.springboard.com
ULTIMATE GUIDE TO DATA SCIENCE INTERVIEWS
76
runawayfromgenericdescriptorsandcompaniesthatstruggletodefinetheircultureorwavethat
questionaway.Greatcompaniesinvesttonsoftimeandeffortintomakingsuretheyhave
awesomepeoplewholovewhattheydo.That’llcomeoffinyourquestioning.
YoushouldalsocheckexternalandobjectivesourcessuchascompanyreviewsonGlassdoor.
ApproachcurrentemployeesaswellasformeronesthatyoucanfindonLinkedIntogettheirside
ofthestory.You’lloftenfindcandidtalesthatcangiveyouagoodpreviewofworkingatyournew
jobwouldbelike.
Team
Companycultureisanextensionoftheteamthatinhabitsit,butyoushouldbeexcitedabout
comingtoworkeverydayandworkingwitheverybodyelse.Makesurethatyou’reworkingwitha
teamthatyoucanlearnfrom.Youarethecombinationofthefivepeopleyouspendthemosttime
with,andyou’regoingtobespendingalotoftimewithyourofficeteam.
Location
Makesurethatyou’recomfortablewherethecompanyislocated,especiallyifyou’removing
significantdistances.Youcan’tmovewithoutgreatdifficulty,andit’simportantthatyoufeelat
easewithwhereyoulive.Mattersliketheweatherandthetransitsystemmattertoacertain
degree,especiallyifyou’regoingtolivewiththoseconditionsforyears.
Negotiating Your Salary
Anastonishing18%ofpeoplenevernegotiatetheirsalary,despitethefactthatthosewhodo
typicallyseetheirsalaryraisedby7%.
Whenyoufirstgetyouroffer,you’reatanuniqueleveragepointthatyoumightnotseeagainfor
severalyears.Thisisthetimetotestwhatyou’reworth.Reachoutwithanofferacompanywon’t
fireyouorcancelacontractofferbecauseyouwereassertingyourworth.Initialoffersaresent
withabufferforslightnegotiation.Takeadvantage.
www.springboard.com
ULTIMATE GUIDE TO DATA SCIENCE INTERVIEWS
77
Duringasalarynegotiation,
1) Comewithawellresearchednumberforwhatyouwant.Looktoindustryaverages,andget
asensefrompeopleworkinginthefieldwhatyoushouldexpect.Nevercomeintoa
negotiationwithoutk
2) Knowingwhatyouwantoutofit.
3) Staypositiveanddon’tpushhardforwhatyou“deserve.”Instead,usethisasapositive
experiencetoassertyourworthandthevalueyoucancreate.
4) Negotiatealittlebithigherthanwhatyouthinkyou’llactuallyget.Anybodyexperiencedat
negotiationwillcomebacktoyouwithacounteroffer,andyou’dbestbepreparedforit.
5) Mostimportantly,don’tfearrejection!Solongasyoukeeptheprocessmovingforward
civillyandprofessionally,acompanywillappreciateyoubeingfrankandpositiveatwhatis
oftenthemostdifficultpartoftherecruitmentprocessforthem.
Beforeyouaccepttheoffer,makesureyouknowhowcommittedyouaretothecompany,team,
andmoney.
 
www.springboard.com
ULTIMATE GUIDE TO DATA SCIENCE INTERVIEWS
78
Facts and Figures
Negotiationisalwayseasierifyouhavesomeaveragesalariestogroundyou.Ifyouhavespecific
offerstopropose,you’llbestrongeratthenegotiationtable.
Herearesomefactsandfiguresthatcanstartyourresearch.
Indeed.comcitesanaveragesalaryof$65,000fordataanalysts,anaveragesalaryof$100,000for
dataengineers,andanaveragesalaryof$115,000fordatascientists.Thisvariesfromregionto
region,withthehighestsalariestendingtoclusterinthetechheavyBayArea.Californiahasthe
highestrangeandmedianofallregionswhenitcomesdatascienceaccordingtoO’ReillyMedia.
Globally,theUnitedStateshasthehighestmedianandrangeofdatasciencesalaries,whilethe
UK,NewZealand,Australia,andCanadaaren’tfarbehind.AsiaandAfricatendtohavethelowest
medians.
Thehighestpayingindustriestendtobetechnologyandsocialnetworkingcompanies,whilethe
lowestpayingonestendtobeeducationandnonforprofitsectors.
Thissalaryalsovariesbasedonskillsandtoolsused.O’Reillyhasadefinitivesurveyofhundreds
ofrespondentsintheindustry.Anopenstudy,theresultsindicateavarietyofdifferentfactors
thatleadtodifferentaveragesalaries.Justasanexample,peoplewhousetheScalalanguage
extensively,aspecializedtypeofprogramming,receiveabove$100,000inmediansalary,while
thosewhouseSPSS,aproprietarytool,earnsignificantlyless.
TakingtheOffertotheBestFirstDay
Ifyou’veacceptedanoffer,congratulations!You’veaccomplishedthegoalofthiswholeprocess
andbrokenintothejobyou’vesought,ajobthatpromisesgoodcompensationandtheabilityto
drivesignificantsocialimpact.
www.springboard.com
ULTIMATE GUIDE TO DATA SCIENCE INTERVIEWS
79
You’llhavetokeepthatmomentumgoingforwardifyouwanttolearnasmuchasyoucan.Be
awarethatcompanieswillworktomakeyouascomfortableaspossible.Youshouldreachoutto
futureteammatesandfigureoutwhotheyareandhowyoucanhelpwiththeirproblemsatwork.
Takethetimetosocializeandmeetasmanypeopleasyoucan.
Moreimportantly,ifyouhavetimebetweenwhenyouacceptedtheofferandwhenyoustart,relax
andenjoy!Makesureyoucatchupwithasmanypeopleasyoucaninyourlife,takethechanceto
rest,andbecompletelyrefreshedforyourfirstdayatwork.
Conclusion
Thedatascienceinterviewprocessisoneofthehardestrecruitmentprocessestocrack,andit’s
oneofthemostcompetitive.Yourfellowintervieweeswillbeadvanceddegreeholders,andsome
ofthemwillhaveextensiveexperienceindatascience.
Whilethefieldisattractingmanytalentedpeople,rememberthatithasaslewofdifferent
industries,challenges,andteamstoworkwith.Ifyouthinkoutsideoftheboxandapplyafew
battletestedtactics,you’llbeabletogetaninterviewandtakeitallthewaytoanofferyoulove.
Splittheprocessintoitscompositesteps,andrememberwhatittakestosucceed.Don’tsearchfor
jobslikeeverybodyelsebyapplyingtothestandardjobpostsandsendingoutforlorncoverletters.
Beinnovativeandsolvecompanyproblemsproactively.Reachouttopeoplewithinthe
organizationforinformationinterviews.Dosomethingdifferentfromthehundredsofother
candidates,andstandoutasagreattechnicalthinkerand,aboveallelse,aproficient
communicator.
Gothroughthetechnicalandnontechnicalpartsofthedatascienceinterview.Onceyou’ve
masteredthethinkingbehindthequestionsandwhathiringmanagersarelookingfor,you’llhave
agoodsenseofhowtoexcelthroughouttheprocess.
www.springboard.com
ULTIMATE GUIDE TO DATA SCIENCE INTERVIEWS
80
Finally,whenyouhaveanoffer(orseveral)onthetable,takethetimetoevaluatethemwithgood
judgement.Takethetimeafteryouacceptanoffertorelax,skillup,andbringthemomentum
forwardtoyourfirstdayatadatasciencejob.
FinalThoughts
“Most of the world will make decisions by either guessing or using their gut. They will be
either lucky or wrong.”- Suhail Doshi
, CEO,
Mixpanel
“The whole enterprise of teaching managers is steeped in the ethic of data-driven analytical
support. The problem is, the data is only available about the past. So the way we’ve taught
managers to make decisions and consultants to analyze problems condemns them to taking
action when it’s too late.”- Clayton M. Christensen
, management professor at Harvard
“We’re entering a new world in which data may be more important than software.”- Tim
O’Reilly
, Founder,
O’Reilly Media
“Web users ultimately want to get at data quickly and easily. They don’t care as much about
attractive sites and pretty design.”- Tim Berners-Lee
“Data scientists are involved with gathering data, massaging it into a tractable form, making it
tell its story, and presenting that story to others.” – Mike Loukides
, VP, O’Reilly Media
Checklist
1) Mapouttheroleyourskillsfit
2) Mapouttheindustriesandtypesofcompaniesyouwanttoworkfor
3) PrepareyourLinkedIn,CV,andemailtemplates
4) Researcheachcompanyandroleyouwanttoaimforthoroughly
5) Reachoutproactivelytoindividualswithincompanieswithinformationalinterviews
6) Buildstrongnetworksandreferrals
7) Tacklethedatascienceinterview
8) Keepuphope
www.springboard.com
ULTIMATE GUIDE TO DATA SCIENCE INTERVIEWS
81
9) Negotiateyouroffer
Templates
Gettinganinformationalinterview
Hi[firstname],
IwassuperinterestedintheproblemsAirBnBisfacingindatascience.I’vebeenaspiringto
breakintothefield,andbeingapassionatefollowerofthe
AirBnBNerds
blog,Inoticedthat
buildingtrustwithdata
isanimportantpartofwhatdrivesAirBnB.Basedonmybackgroundin
psychologyandstatistics,Imightbeabletohelpcomeupwithsomecreativeideasonhowto
fostertrust.I’dlovetotakeyououttocoffeeandgetagreatersenseofwhatproblemsAirBnB
hasperhapsIcanhelp!
Cheers,
[yourname]

[Greeting],
[Whyareyouinterestedinthecompany],[somethingthecompanyhasdonethatyoulove],[how
youcanhelp].
 
www.springboard.com
ULTIMATE GUIDE TO DATA SCIENCE INTERVIEWS
82
Reachingouttogetareferral
Hi[firstname],
Itwasgreatseeingyouatthepotluck!I’vebeenlookingaround,andI’minterestedinthe
problemsUberisfacing,specificallytheonesfacedbydatascientistsonthegrowthteam.Would
youmindintroducingmetothehiringmanagerorsomebodyontheteamsoIcouldseeifIcould
help?
Cheers,
[yourname]

[Greeting],
[Talkaboutlastpointofcontact],[talkaboutinterestincompanyandproblemsfacedbya
specificrole],[asktobeintroducedtohiringmanagertohelpsolvethoseproblems]
Followingupafteraninterview
Hi[askhowyourinterviewerpreferstobeaddressed],
ItwasapleasuretalkingwithyouaboutGoogle’sdatascienceproblems.IthinkIcanhelpwith
someoftheproblemsyou’veenumerated,andIlookforwardtothenextstepsintheprocess!

Hello[Askyourinterviewerhowtheyprefertobeaddressedduringtheinterview],
[Talkaboutproblemsyoucanhelpsolve],[Statethatyou’relookingforwardtonextsteps]
www.springboard.com
ULTIMATE GUIDE TO DATA SCIENCE INTERVIEWS
83
Glossary
A/BsplittestAnA/Bsplittestisthegoldenstandardofexperimentdesignforwebcompanies,
wheretwogroupsofusersaresubjectedtodifferenttreatmentsandmeasuredtoseetheir
conversionratetoacertaingoal.Optimizely,awebcompanydedicatedtohelpingrunA/Bsplit
testshasagoodguideontheconcept.
FeatureAnuggetofinformationaboutanobject,usuallystoredasacolumnintabulardata.If
youmeasureandstoretheheight,weight,andgenderofanindividual,youarestoringthree
featuresaboutthem.
LifetimeValueTheexpectedamountofrevenueacustomerisexpectedtogenerateoverthe
timetheyspendwithacompany.Asoftwareasaservicestartupthatsellssoftwarebythemonth
canexpecttocalculatethisbymultiplyingthemonthlypricewiththenumberofmonthsspent.
MapReduceAsetofalgorithmsthatacttoabstractawaythedifficultyofstoringmassivedata
setsbytreatingdatasplitintomultipleserversinawayasintuitiveashandlingitfromone.
MapReduceusesparallel,distributedlogictodealwithmassivedatasets.
OverfittingThetendencyofamodeltofitontopastdata,overgeneralizingfromthoseinsights
tomakeinaccuratepredictionsinthefuture,draggeddownbytheweightofthepast.
TypeIErrorAfalsepositiveistheincorrectacceptancethatsomethingishappeningakinto
tellingamanthatheispregnant.Intechnicalterms,itistheincorrectrejectionofthenull
hypothesis.
TypeIIErrorAfalsenegativeistheincorrectacceptancethatsomethingisn’thappening.Itis
akintotellingapregnantwomansheisn’tpregnant.Intechnicalterms,itistheincorrect
acceptanceofthenullhypothesis.
Formoreglossaryterms,consultthisdatascienceglossary.
www.springboard.com
ULTIMATE GUIDE TO DATA SCIENCE INTERVIEWS
84
Resources
AparodyoftheinterviewprocessthatexaminessomehardtruthsfromKDNuggets.
Thisbook,calledDataScienceInterviewsExposed,offersmoresamplequestionsthatyoucan
tacklewithyourinterviewpractice.
TheDataScienceHandbookoffersreallifeadvicefromdatascientists,includingsomesmart
analysisonwhatmakesforagreatdatascientistandwhathappensduringtheinterviewprocessto
findthoseindividuals.Itscompanion,theDataScienceInterviewGuide,offers120questionsyou
mightseeinadatascienceinterview.
CrackingtheCodingInterviewisadefinitiveresourceforgoingthroughsoftwareengineering
interviewsandwillhelpwiththeprogrammingpartsofthedatascienceinterview.
ThisQuorathreadgoesintohowAirBnBhiresfordatascientists,aninsightfullookatthedata
scienceinterviewprocessfromanestablisheddatascienceleader.
ThisexposebyTreyCauseyexplainshowtoacethedatascienceinterviewprocessandoffersa
criticalandunvarnishedlookonhowoneshouldapproachthedatascienceinterview.Erin
Shellmanalsotalksaboutherexperiencegettingajobindatascience.
“AsI'vegottenolderandmoreexperienced,Ipushbackininterviews.Iaskquestionsaboutwhat
thepurposeofaproblemisorstatethatIdon'tthinkthisisagoodevaluationofmyskillsor
abilities.SomepeopleprobablyseethisasmethinkingI’m"toogood"toanswerthequestions
everyoneelsehastoanswer,butIseeitasdoingmyparttobeacriticalthinkerabout
evaluation,prediction,andhiring.Hopefullyyou'lldothistooandasmoreofusareinaposition
wherewearebuildingteamsandhiring,we'llthinkmorecarefullyaboutwhatwe'retryingto
accomplishandhowwecangetthereinsteadofcopyingthesamepatternsthathavebeen
aroundforyears.
”
ThisarticleisaninsightfulreadabouthowdatascienceatTwitterworksandofferstheinside
perspectiveofsomebodywhoisadatascientistintheindustry.
Ifyoufindyourselfthinkingaboutprobability,refertothischeatsheettomakesureyou’reontop
ofanyproblem.ThisQuorathreadwillhelpaswell.
EllenChisawritesaboutthingsshehasscreweduponwhenitcomestotechnicalinterviews;you
shouldmakesuretoavoidthosemistakes!
www.springboard.com
ULTIMATE GUIDE TO DATA SCIENCE INTERVIEWS
85
Finally,FirstRoundReviewhasaprimeronhowtohireexceptionaldatascientists;readitto
knowhowthepeopleontheothersideofthetablethink.
AbouttheCoAuthors
Rogerhasalwaysbeeninspiredtolearnmore.Hebrokeintoacareerindatabyanalyzing$700m
worthofsalesforamajorpharmaceuticalcompany.HehaswrittenforEntrepreneur,
TechCrunch,TheNextWeb,VentureBeat,andTechvibes.
Forthisguide,hecompiledinsightfromSpringboard'snetworkofhundredsofdatascience
experts,includingSriKanajan,hiscoauthor.
SriKanajaniscurrentlyaseniordatascientistinNewYorkCityatamajorinvestmentbank.He
has14yearsofexperienceinvariousengineeringandmanagementcapacitiesandmadeacareer
transitiontobeadatascientistin2013.HecompletedafulltimedatasciencebootcampinSan
Franciscoandprogressedtobecomeadatascientistattwostartupsandeventuallyadatascience
manageratChange.orgbeforetakingonhiscurrentrole.Srialsoteachesparttimeasalead
instructorinGeneralAssembly'sDataSciencecourse.Heispassionateabouthelpingothersmake
thetransitionintodatascience.
www.springboard.com
ULTIMATE GUIDE TO DATA SCIENCE INTERVIEWS
86

Navigation menu