← All categoriesMLOps & Production
What happens after the notebook. Pipelines, monitoring, drift and rollbacks.
Scroll this category (28) →Ready to read
28🟡Data Drift vs Concept Driftmust-know4 min🟡What to Monitor in Productionmust-know5 min🔴Point-in-Time Correct Feature Joinsmust-know5 min🔴Debugging a Production Model Incidentmust-know5 min🟢The End-to-End ML Lifecycle5 min🟢Experiment Tracking & Reproducibility5 min🟢Model Registries & Promotion5 min🟢Model Cards & Documentation5 min🟡Feature Stores5 min🟡Model & Data Versioning5 min🟡CI/CD for ML5 min🟡Testing ML Code & Data5 min🟡Shadow & Canary Deployments5 min🟡Rollbacks & Kill Switches5 min🟡Model Serving Patterns5 min🟡When (and How Often) to Retrain5 min🟡Docker & Kubernetes for ML5 min🟡Orchestration: Airflow, Dagster, Prefect5 min🟡SLAs, SLOs & Error Budgets for ML5 min🟡Attributing & Cutting ML Spend5 min🟡Edge vs Cloud Inference5 min🔴Detecting Drift: PSI, KS, KL5 min🔴Monitoring When Labels Arrive Late5 min🔴Autoscaling Inference Workloads5 min🔴GPU Utilization & Cost Control5 min🔴ONNX, TensorRT & Runtime Export5 min🔴vLLM, TGI & LLM Serving Stacks5 min🔴Model Supply-Chain Security5 min