About the role
Requirements:
- CI/CD for ML: pipelines, testing, packaging.
- Model registry, experiment tracking (MLflow or similar).
- Serving: REST/gRPC, batch vs online, latency/throughput tradeoffs.
- Monitoring: data drift, model drift, performance, alerts.
- Governance: lineage, reproducibility, approvals.
Good-to-have:
- Feature store, automated retraining strategies.
- Kubernetes, Terraform/IaC.
- Canary/shadow deployments, A/B testing for models.