// the find
DataTalksClub/mlops-zoomcamp
Free MLOps course from DataTalks.Club
A free 9-week course teaching the MLOps lifecycle end-to-end: experiment tracking with MLflow, orchestration with Prefect, deployment via Flask/Lambda/batch, and monitoring with Evidently and Grafana. Aimed at data scientists and ML engineers who know the modeling side but haven't shipped a production ML system. The course has been running since 2022 with multiple cohorts, so the material is well-worn.
Each module pairs a concept with a specific tool and working code — you're not just reading theory, you're running MLflow against a real model and a real dataset (NYC taxi trips). The monitoring module is better than most courses: it wires Evidently data drift reports into Grafana dashboards via Prometheus rather than stopping at 'run a report in a notebook'. The best-practices module covers Terraform IaC and GitHub Actions CI/CD in a concrete AWS pipeline, not just linting tips. The cohort archive (2022–present) means you can see how homework problems and solutions have evolved as tools changed.
No live cohort planned for 2026, so you lose peer review and graded feedback — the self-paced path is just you and a Slack channel. The orchestration module (Module 3) is noticeably thinner than the others: the README exists but the content relies heavily on video, which means the repo alone doesn't stand on its own. AWS-centric throughout — Kinesis, Lambda, ECR, S3 — so if you're on GCP or Azure you'll be translating the whole deployment and IaC sections yourself. The monitoring toolchain (Evidently + Grafana + MongoDB + Prefect) is a lot of moving parts to stand up locally, and the post-evidently-0.7 fork in Module 5 signals the content is chasing a fast-moving library.