Maintain and evolve an existing recommendation system in a multi-tenant environment
Bring models from experimentation into production and improve their real-world performance
Build and support end-to-end ML pipelines : training, validation, deployment, retraining
Optimize training/inference workloads for latency, reliability, memory usage, and scalability
Integrate ML inference into Python-based backend services and collaborate with backend/data teams
Define and track model KPIs linked to product and business impact (engagement/retention/revenue)
Improve data quality , feature availability, and feature pipelines (Airflow-based workflows)
Set up monitoring for model health: drift, degradation, anomalies, incident response
Run controlled experiments (A/B tests), analyze results, and translate insights into improvements
Document model behavior, assumptions, and operational runbooks
Contribute to architecture decisions for scalable ML infrastructure and deployment practices

5+ years in ML Engineering / Production ML roles
Degree in a quantitative field (Math/Stats/CS or similar)
Strong Python skills and experience building production-grade ML services
Solid ML foundation: supervised learning, ranking/recommendations, evaluation methodology
Hands-on experience with feature engineering for event/behavioral data
Production deployment experience: APIs and/or batch jobs, Airflow, CI/CD, containers
Practical SQL knowledge and understanding of data access patterns
Experience with monitoring ML systems end-to-end: data quality + model performance + alerts
Understanding of experimentation and statistics (A/B testing, experiment design)
Strong engineering habits: testing, code review, documentation
Computer science fundamentals (processes, memory, performance considerations)

ML Engineer

Key skills