MLOps In Practice – How To Run Your Machine Learning Models In Production At Enterprise Scale
summary
Shows a practical reference architecture: a batch training pipeline that writes versioned model artifacts to a model registry and a separate real‑time scoring path that serves models behind an API or embeds them in stream processors (Flink/Spark/Kafka) when latency demands it. Recommends treating models like code with CI/CD pipelines, collecting actuals for feedback, using a feature store to avoid expensive runtime feature computation, and choosing between AutoML, vendor stacks or custom models based on data uniqueness.
AI-written summary · about
chapters (11)
- 0:00Introduction and speaker intro
- 1:03Basics of machine learning in production
- 3:54MLOps as cultural and team shift
- 7:28Roles in multi-disciplinary ML teams
- 9:15Best practices for ML model development
- 14:10Reference architecture for ML systems
- 20:27CI/CD and deployment for ML models
- 22:35Tooling considerations and trade-offs
- 26:10Model serving options and approaches
- 29:19Orchestration and workflow tools
- 30:00Wrap-up and key takeaways
related talks
41mTHE STATE OF MLOPS - MACHINE LEARNING IN PRODUCTION AT ENTERPRISE SCALE
10mFrom Model to Production: Building Scalable MLOps Pipelines
47mMachine Learning in Production: An Intro to MLOps.
41mFrom Prototype to Production Building a Scalable MLOps Platform from a Reference Arch
45mThe Culture, Architecture and Technology of MLOps
53mExperience with an end-to-end MLOps pipeline: Practical insights and 3D architectural perspectives
25mBuilding Scalable MLOps Pipelines That Actually Work
13mMLOps: From AI Research to Production-Ready Systems