• Design, train, and evaluate ML models for logistics use cases: delivery time estimation, demand forecasting, capacity planning, anomaly detection • Improve and iterate on existing production models using performance data and customer feedback • Run structured experiments to validate model improvements before promotion to production • Own the full path from trained model to production API — wrapping, deploying, versioning, and monitoring • Build and maintain inference APIs serving predictions at scale (<100ms latency targets) • Deploy and manage models on GCP Vertex AI • Implement A/B testing and rollback strategies for safe model promotion • Build real-time and batch feature pipelines from Postgres/AlloyDB sources • Design feature stores serving both training and inference • Implement data validation and quality monitoring to catch drift before it affects customers • Develop CI/CD pipelines for model deployment • Monitor model and pipeline health; own incident response for ML systems • Optimize inference costs across GCP and Cloudflare infrastructure • Partner with the Data Scientist on experiment design and feature strategy • Work with platform engineers to integrate ML outputs into core product services • Communicate model behavior, limitations, and tradeoffs clearly to non-ML engineers and product stakeholders