MLOps Transformation
summary
Ahold Delhaize runs a per-project ML maturity assessment and moves brands from Level 0–2 toward Level 3 by enforcing four MLOps standards—documentation, code quality, traceability, monitoring—and providing a golden path (cookie‑cutter repos, CI/CD, templates). Implementation uses GitHub Actions + Databricks for training IDs, MLflow model registry, Docker images that carry git commit hash and run ID, and centralized monitoring; it takes about a year to change ways of working and enables data teams to deploy and rollback models reliably.
AI-written summary · about
chapters (10)
- 0:00Introduction to MLOps
- 1:27MLOps adoption challenges
- 3:34Organization goals and maturity
- 5:42Maturity assessment example
- 7:50Wall of confusion in teams
- 8:54Breaking the wall with standards
- 12:43Changing team responsibilities
- 13:25Three-part transformation recipe
- 13:46Golden path for MLOps
- 17:38Key takeaways and Q&A
related talks
43mNavigating the MLOps Journey: Key Considerations for Successful Implementations
28mMastering MLOps: Building Future Ready ML Systems
01mMLOps: Applying DevOps to ML
53mExperience with an end-to-end MLOps pipeline: Practical insights and 3D architectural perspectives
43mImplementing MLOps practices
13mMLOps at Scale
08mMLOps in Healthcare: Scaling AI Model Deployment
22mWhy you need to think about MLOps at the beginning of your project