Machine Learning Engineer

Job DescriptionJob OverviewThe Data Scientist / ML Engineer builds and deploys predictive models and analytical systems that turn AGS's player and game data into quantitative insights that directly improve game design and commercial decisions. This role bridges behavioral data science (understanding how players interact with games) and production ML engineering (deploying models that actually reach decision-makers). It feeds game designers with data-driven design recommendations for the ML-driven game design initiative, supports yield management with predictive models for Interactive YieldMax, and enables operators to understand their player base more deeply — anchored to AGS's Tech & Data hero mission of an accessible data layer with live KPIs powering every decision.ResponsibilitiesBuild player session behavioral models — retention prediction, abandonment modeling, post-bonus behavior analysis, and bet escalation modeling from iGaming session dataDevelop game performance prediction models — predict WPUPD, time on device, and floor longevity from game specification features and historical performance data, using a game feature extraction pipeline that reverse-engineers existing titles into structured, reusable featuresBuild math model optimization analytics — analyze actual vs. theoretical RTP, hit frequency, and bonus frequency; identify math model anomalies across the deployed fleetCreate player segmentation models — cluster players into behavioral archetypes (bonus hunters, jackpot chasers, base game grinders) to inform game design and operator recommendationsSupport the Interactive YieldMax yield-management tool — build the underlying models that predict which AGS game maximizes performance in a given floor position, operator property, and player demographicBuild predictive maintenance models — analyze cabinet error logs and, as sensor/telemetry pipelines mature (Dynamics Field Service / Dataverse), incorporate telemetry to identify failure precursor patterns and predict component failuresFeed game design decisions — translate model outputs into game designer-friendly insights that are actionable in the game specification processDesign and analyze A/B tests — experimental design, statistical analysis, and results interpretation for game math variant testing (where regulatorily permitted)Productionalize models — package models for deployment on Azure ML/Fabric, with MLflow-based registry, monitoring, and retraining pipelinesSkills/Requirements4–8 years of data science and/or ML engineering experience, with demonstrated production model deployment (not just notebook analysis)Behavioral analytics expertise — has built retention, churn, or engagement models using event-level behavioral data (session logs, clickstreams, transaction sequences)Strong Python and SQL skills — pandas, scikit-learn, XGBoost, statsmodels; can query the data warehouse independently (a mix of on-prem SQL Server and Salesforce today, migrating to Microsoft Fabric/OneLake) without relying on a data engineer for every analysisStatistical rigor — survival analysis, A/B test design, causal inference, regression modeling; understands the difference between correlation and causationMachine learning breadth — classification, regression, clustering, recommendation systems; can select the right modeling approach for each problemData communication skills — can translate model outputs into business-friendly language that game designers and commercial leaders can act onExperience with messy, real-world data — comfortable where game features aren't fully documented and pipelines are still being built; doesn't require perfect data to deliver valueBachelor's or Master's degree in Data Science, Statistics, Computer Science, Mathematics, or related quantitative fieldPreferredGaming, mobile gaming, or consumer behavioral analytics experienceFamiliarity with casino game mechanics — RTP, volatility, Hold & Spin, theo indexExperience with time series analysis and anomaly detection for IoT/sensor dataKnowledge of responsible gambling data considerationsExperience with MLflow, Azure ML, or Fabric Notebooks/Spark for model lifecycle managementNote: All offers are contingent upon successful completion of a background check*Posted positions are not open to third party recruiters and unsolicited resume submissions will be considered free referrals.AGS is an equal opportunity employer