Workshop
Deploying AI at Scale: MLOps for Production Teams
Your model works in the notebook. It doesn't survive production.
About this workshop
The engineering playbook for taking ML models from notebook to production. Covers CI/CD for ML, model registries, monitoring for drift, A/B testing AI features, and infrastructure patterns used at Google scale.
What you will learn
- Build a CI/CD pipeline for ML that automatically tests, versions, and deploys models
- Set up a model registry and manage model lineage across environments
- Implement data drift and concept drift monitoring with automated alerting
- Run statistically valid A/B tests for AI features with proper guardrails
Who this is for
- ML engineers whose models work in notebooks but struggle to survive production
- Platform and infrastructure engineers taking ownership of ML deployment
- Engineering leads building the MLOps foundation for a growing AI team
By the end
Before
Models that degrade silently after deployment
After
Drift monitoring with automated alerting before users ever notice
Before
Manual, error-prone deployment processes
After
A CI/CD pipeline that tests, versions, and deploys automatically
Before
A/B testing AI features with no statistical rigour
After
Properly designed experiments with clear guardrails and confidence
About Carlos
Carlos Reyes
MLOps Lead, Google DeepMind Alumni
Vetted by Maram
Carlos led MLOps infrastructure at Google DeepMind, where he built CI/CD pipelines for large-scale model deployment. He now consults for high-growth AI companies and is a co-author of the O'Reilly report on production ML systems.
View full profile →What learners say
Reviews appear here once 3 learners have completed this session.