MLOps, governance and trust
summary
Cites 90% McKinsey and survey numbers: 84% of data scientists distrust their models, 70% of executives see no AI value. Describes DataRobot MLOps: agents register models in a model registry, monitor accuracy and data drift, support Python/Java/R, deploy to AWS/GCP/Azure or Kubernetes, and run champion/challenger tests. Shows model hot swapping with zero interruption, troubleshooting a 10%+ accuracy drop, auto-rules to override missing feature values, and a retail case with 8x precision and $20–30M ROI.
AI-written summary · about
chapters (9)
- 0:00Intro
- 0:35Data About
- 2:25Silo
- 9:45Notifications
- 14:40Troubleshooting
- 19:50Model Registry
- 25:15Hot Swapping
- 26:02Risk Management
- 27:25Use Cases
related talks
13mMLOps at Scale
26mAre you ready for MLOps? 🫵
30mData Prep Ops: The Missing Piece of the MLOps Puzzle
14mGetting Machine Learning applications into production using MLOps and CD4ML
53mExperience with an end-to-end MLOps pipeline: Practical insights and 3D architectural perspectives
13mMLOps: From AI Research to Production-Ready Systems
47mMachine Learning in Production: An Intro to MLOps.
59mInside the fastest-growing trends in MLOps: VC & startup perspective