# CI/CD for machine learning A pack of 8 sessions from the MLOps Community YouTube channel, in the order to watch them. 7h 26m of video. Page: https://mlopstalks.com/packs/ci-cd-for-machine-learning A rollback restores yesterday's model weights behind today's incompatible serving code. Another release leaves the preprocessing behind. Preventing these failures means treating model deployment as a versioned change: record the trained artifacts, transformations, configuration and runtime together, then test the package that production will run. Start by mapping that path into the application's release process. Extend familiar CI practices to data and evaluation reports, then distinguish releasing a pipeline from releasing the model it produces. Packaging and environment promotion make the release concrete. Shadow traffic provides evidence before predictions affect customers, and explicit deployment contracts decide what may advance. Finish by keeping the release path exercised between launches. These 2020-2022 accounts explain delivery decisions; their historical tool demonstrations are examples, not current setup instructions. ## This pack is for you if - Training finishes automatically, but deploying its model still requires a list of manual steps. - A rollback restores the old weights but leaves incompatible preprocessing or serving code deployed. - Your release pipeline worked last year, and nobody knows whether it still runs. ## The talks, in order ### 1. CI/CD in MLOps Monmayuri Ray, GitLab | 50:25 | MLOps Coffee Sessions Video: https://www.youtube.com/watch?v=5EM211r3Xaw Summary: https://mlopstalks.com/talks/ci-cd-in-mlops.md Why first: Ray puts the model inside the application's delivery process. A commit may trigger training, registration and scoring before an accepted result moves to deployment. Mapping those steps first makes it easier to decide what the existing CI system can support and where manual decisions remain. ### 2. Continuous Integration for ML Elle O'Brien, Iterative | 1:01:46 | MLOps Coffee Sessions Video: https://www.youtube.com/watch?v=L98VxJDHXMM Summary: https://mlopstalks.com/talks/continuous-integration-for-ml.md Why second: O'Brien extends the familiar Git workflow to the inputs and evidence that ML changes require. Dataset changes must be visible to CI, and reviewers need plots and metrics alongside a pass or fail. Her 2020 examples explain what to attach to a proposed release before automating the rest of its journey. ### 3. CI/CD & Continuous Training in ML David Hershey, Determined AI | 1:00:53 | MLOps Coffee Sessions Video: https://www.youtube.com/watch?v=DIwtPQN9cus Summary: https://mlopstalks.com/talks/ci-cd-continuous-training-in-ml.md Why here: The artifact being delivered is not always the trained model. Hershey separates building pipeline components, deploying the pipeline, running it and publishing its model output. That distinction prevents one successful job from standing in for the whole release, and makes the required records easier to specify. ### 4. Building Machine Learning Models into Docker Images Luke Marsden, MLOps Consulting | 1:06:27 | MLOps Meetup Video: https://www.youtube.com/watch?v=CHttwWGdWK4 Summary: https://mlopstalks.com/talks/building-machine-learning-models-into-docker-images.md Why here: Marsden asks exactly what the release contains. His preference for baking weights into a serving image makes the model and runtime one versioned unit; downloading weights instead leaves two versions and a startup dependency to track. This is the packaging decision to settle before promoting anything between environments. ### 5. Deploying Machine Learning Models at Scale in Cloud Vishnu Prathish, Innovyze | 57:53 | MLOps Meetup Video: https://www.youtube.com/watch?v=i3U0gkHX24s Summary: https://mlopstalks.com/talks/deploying-machine-learning-models-at-scale-in-cloud.md Why here: Prathish adds the transformations that a model file can leave behind. Innovyze recorded cleaning and feature engineering as configuration and used a shared path from development through QA to production for customer-specific models. Packaging is only useful when the promoted result still receives the inputs it was built for. ### 6. The Shipyard: Lessons Learned While Building an ML Platform Joseph Haaga, Interos | 40:00 | MLOps Coffee Sessions Video: https://www.youtube.com/watch?v=_n-GthGQw7o Summary: https://mlopstalks.com/talks/the-shipyard-lessons-learned-while-building-an-ml-platform.md Why here: Haaga makes promotion a recorded change. Interos checked the expected project structure automatically, then changed image tags and model-weight paths in versioned environment repositories. The release can be reviewed as a specific change instead of reconstructed from someone remembering which commands they ran. ### 7. Monzo Bank - An MLOps Case Study Neal Lathia, Monzo Bank | 1:03:54 | MLOps Coffee Sessions Video: https://www.youtube.com/watch?v=EyLGKmPAZLY Summary: https://mlopstalks.com/talks/monzo-bank-an-mlops-case-study.md Why here: Lathia separates running a new model from acting on its predictions. Monzo's shadow mode logs real requests and outputs while customers remain on the existing behavior. Its different validation requirements for different risks give promotion a decision point beyond whether the deployment completed. ### 8. MLOps at Stripe Emmanuel Ameisen, Stripe | 44:36 | MLOps Coffee Sessions Video: https://www.youtube.com/watch?v=L8WQBYCRaGc Summary: https://mlopstalks.com/talks/mlops-at-stripe.md Why last: Stripe turned business expectations into release contracts, then kept data generation, training and evaluation running even without a planned launch. Ameisen's warning is practical: a delivery path that sits unused decays. Automation earns trust by repeatedly exercising the route you will need for the next release. ## Editor's note Hershey separates pipeline releases from model outputs, while Ameisen keeps the producing and checking jobs running between launches. ZenML records each pipeline run's steps, inputs, outputs and code version, and lets the same pipeline run on different infrastructure through configuration. Those records help connect a release artifact to the workflow that produced it; the team's deployment contracts still determine whether it should advance. Written by the MLOps Talks editors (the ZenML team), not by any of the speakers.