Please note: this role is for our partner company, a game studio developing slot and casino games with a strong focus on data-driven growth and AI-powered engagement solutions. The OpportunityWe are Art of Retention, a new standalone product unit backed by a major iGaming holding (Altenar/Agregain). We are building the "Brain" that predicts player behavior (Churn, LTV, VIP) and automates retention actions for millions of global users.We have the Data Scientist to build the math. We need you to build the factory. This is a Greenfield opportunity: no legacy code, no technical debt. You will design the inference infrastructure from scratch, transitioning our stack from "Phase 1 Offline Batch" to "Phase 2 Real-Time Production."️ The MissionYour mandate is operational excellence. You will own the model lifecycle from the moment the Data Scientist commits a script to the moment the API returns a probability score to the operator.You don't just train models; you ensure they run reliably, scale automatically, and retrain when the data changes. Key ResponsibilitiesModel Deployment: Wrap XGBoost/LightGBM models into high-performance APIs (FastAPI/Docker) capable of handling high-load requests from Sportsbook platforms.The "Retraining Loop": Build automated pipelines. When data drifts, the system should trigger a retrain automatically—eliminating manual script runs.Infrastructure & Registry: Deploy a Model Registry (MLflow) and Feature Store. Ensure we can rollback to "Model v1.2" instantly if "Model v2.0" degrades.Lean Architecture: Design a cost-effective infrastructure on AWS or cloud-agnostic containers. We avoid expensive "Vendor Lock-in" (like heavy SageMaker dependencies) where simple Docker containers work better.Monitoring (Drift): Implement alerting for Data Drift and Model Decay. You are the first to know if the model stops predicting accurately. The StackCore: Python (Expert), SQL.Serving: FastAPI, Docker, Kubernetes (K8s).ML Ops: MLflow, Airflow (or Prefect/Dagster).Cloud: AWS (EC2/EKS/S3) or similar.CI/CD: GitHub Actions / GitLab CI. Who You AreYou are an Engineer first. You write modular, tested, production-grade code, not just Jupyter Notebooks.3+ Years Experience: Specifically in ML Engineering or MLOps.Tabular Data Pro: You have deployed Gradient Boosting models (XGBoost/CatBoost) for transactional data. (We don't need LLM/Computer Vision experts).Container Native: You know how to scale an inference service using Docker/K8s.The "Plumber" Mindset: You enjoy building the pipes that make data flow reliably. Why Join Us?Autonomy: You are the first ML Engineer. You choose the architecture (within reason).Impact: Your code directly drives revenue for major global operators.Stability: We operate with the speed of a startup but are backed by a profitable, established holding.