AI and ML imbedded in Azure Data Platform
summary
Clear guidance on when to build custom models versus calling pre‑trained APIs, with a live walkthrough of Azure Machine Learning Studio to create a workspace, train and export models, and Cognitive Services for quick sentiment, vision and keyphrase tasks (Microsoft lists over thirty services). Covers containerizing models, edge deployment via Azure Stack/Kubernetes for low latency, integration points into Azure Synapse, Power BI and Databricks, and the cost tradeoff: development time versus per‑call billing.
AI-written summary · about
chapters (8)
- 0:00Introduction and speaker intro
- 8:14Speaker background and session overview
- 13:15AI vs ML concepts explained
- 27:48Machine learning vs cognitive services
- 36:57Azure reference architectures for AI
- 43:25Deploying models at the edge
- 52:01AI in Azure Data Platform services
- 1:00:14Resources and closing
related talks
43mA lap around Azure Machine Learning services
30mData & AI platform components in Azure
34mMaking Data Scientists Productive in Azure
53mDemystifying Artificial Intelligence with Microsoft Cloud, Azure!
51mDeveloping Solutions with Azure Machine learning studio
58mDeep Learning in Microsoft Azure: CNTK, CaffeOnSpark and Tensorflow
01h 14mAI of Things
31mAzure AI Innovations at Microsoft