Machine Learning Engineer

The Data Scientist /MLEngineerbuildsanddeployspredictivemodelsandanalyticalsystemsthatturnAGS'splayerandgamedataintoquantitativeinsightsthatdirectlyimprovegamedesignandcommercialdecisions.

Thisrolebridgesbehavioraldatascience(understandinghowplayersinteractwithgames) andproductionMLengineering(deployingmodelsthatactuallyreachdecision-makers).Itfeedsgamedesignerswithdata-drivendesignrecommendationsfortheML-drivengamedesigninitiative,supportsyieldmanagementwithpredictivemodelsfor

Interactive Yield Max ,andenablesoperatorstounderstandtheirplayerbasemoredeeply—anchoredtoAGS'sTech&DataheromissionofanaccessibledatalayerwithliveKPIspoweringeverydecision.

Responsibilities

• Buildplayersessionbehavioralmodels—retention prediction,abandonment modeling,post-bonusbehavioranalysis,andbetescalationmodelingfromiGamingsessiondata
• Developgameperformancepredictionmodels—predictWPUPD,timeondevice,andfloorlongevityfromgamespecificationfeaturesandhistoricalperformancedata,usingagamefeatureextractionpipelinethatreverse-engineersexistingtitlesintostructured,reusable features
• Buildmathmodeloptimizationanalytics—analyzeactualvs.theoreticalRTP,hit frequency,andbonusfrequency;identifymathmodelanomaliesacrossthedeployedfleet
• Createplayersegmentationmodels—clusterplayersintobehavioralarchetypes(bonus hunters,jackpot chasers,basegamegrinders) toinformgamedesignandoperatorrecommendations
• Support the Interactive Yield Maxyield -management tool—buildtheunderlyingmodelsthatpredictwhichAGSgamemaximizesperformanceinagivenfloorposition,operator property,andplayerdemographic
• Buildpredictivemaintenancemodels—analyzecabineterrorlogsand,assensor/telemetrypipelinesmature(Dynamics Field Service /Dataverse),incorporatetelemetrytoidentifyfailureprecursorpatternsandpredictcomponentfailures
• Feedgamedesigndecisions—translatemodeloutputsintogamedesigner-friendlyinsightsthatareactionableinthegamespecificationprocess
• DesignandanalyzeA/Btests—experimental design,statistical analysis,andresultsinterpretationforgamemathvarianttesting(whereregulatorilypermitted)
• Productionalizemodels—packagemodelsfordeploymenton
  
  AzureML/Fabric,withMLflow-based registry,monitoring,andretrainingpipelines

Skills/Requirements

• 4–8 years of data science and/or ML engineering experience
  , with demonstrated production model deployment (not just notebook analysis)
• Behavioralanalyticsexpertise
  —hasbuiltretention,churn,orengagementmodelsusingevent-levelbehavioraldata(session logs,click streams,transaction sequences)
• StrongPythonandSQLskills
  —pandas,scikit-learn,XGBoost,stats models;canquerythedatawarehouseindependently(amixofon-premSQLServerandSalesforcetoday,migratingtoMicrosoftFabric/One Lake) withoutrelyingonadataengineerforeveryanalysis
• Statistical rigor
  —survival analysis,A/Btestdesign,causal inference,regression modeling;understandsthedifferencebetweencorrelationandcausation
• Machinelearningbreadth
  —classification,regression,clustering,recommendation systems;canselecttherightmodelingapproachforeachproblem
• Datacommunicationskills
  —cantranslatemodeloutputsintobusiness-friendlylanguagethatgamedesignersandcommercialleaderscanacton
• Experiencewithmessy,real-worlddata
  —comfortablewheregamefeaturesaren'tfullydocumentedandpipelinesarestillbeingbuilt;doesn'trequireperfectdatatodelivervalue
• Bachelor'sor Master'sdegreein
  
  Data Science,Statistics,Computer Science,Mathematics,orrelatedquantitativefield

Preferred

• Gaming,mobile gaming,orconsumerbehavioralanalyticsexperience
• Familiaritywithcasinogamemechanics—RTP,volatility,Hold&Spin,theoindex
• Experiencewithtimeseriesanalysisandanomalydetectionfor
  
  IoT/sensor data
• Knowledgeofresponsiblegamblingdataconsiderations
• Experience with MLflow, Azure ML, or Fabric Notebooks/Spark for model lifecycle management

Note:

Alloffersarecontingentuponsuccessfulcompletionofabackgroundcheck

AGSisanequalopportunityemployer

Location:

Atlanta,Georgia

Job Type:

Executive/HR/IT

Date Posted:

07/20/2026

AGS is a global company focused on creating a diverse mix of entertaining gaming experiences for every kind of player. Our roots are firmly planted in the Class II Native American gaming market, and our customer-centric culture and growth have helped us branch out to become a leading all-inclusive commercial gaming supplier. Powered by high-performing Class II and Class III slot products, an expansive table products portfolio, real-money gaming platforms and content, highly rated social casino solutions for operators and players, and best-in-class service, we offer an unmatched value proposition for our casino partners.

Learn more at  .