Beginners guide to Azure MLOps
summary
Beginner-level walkthrough of Azure MLOps showing a real engineering-client pipeline: uses Azure Machine Learning workspace, Databricks, Jupyter/VS Code, container registry, Key Vault, Application Insights and Azure Pipelines to create reproducible pipelines, reusable environments, model registration and deployments (including edge via FPGA). Demonstrates a regression price-prediction pipeline that runs successfully in the designer, captures governance/alerts/monitoring, and claims parallel training speedup from 3 hours to 25 minutes for ~200 models while automating retrain and redeploy.
AI-written summary · about
chapters (8)
- 0:00Welcome and introduction
- 1:05Machine learning process
- 7:47Azure ML service overview
- 12:26Azure ML pipelines
- 14:13Experimentation tracking
- 17:23MLOps process explained
- 19:34Demo of Azure ML studio
- 26:44Q&A and closing
related talks
45mMLOps and ML-Flops: going from dev to prod
48mStreamlining your Machine Learning projects with Azure Machine Learning
51mHow to get started with MLOps in Azure Databricks
24mFrom Model to Production: Automating the ML Lifecycle with CI/CD
44mPractical deep dive into machine learning techniques and MLOps
57mDevOps for Machine Learning
52mDevOps for AI in the Microsoft World
10mFrom Model to Production: Building Scalable MLOps Pipelines