Microsoft · AI-200
Cosmos DB, vectors and change feed
AI-200 · Lesson 5 of 13 · 34:31
YouTube loads only after you press play. This chapter opens at its timestamp in the full course. Watch on YouTube ↗
In this lesson
Prepare for the AI-200 exam (Microsoft Certified: Azure AI Cloud Developer Associate). In this lesson: vector embedding policies, flat, quantizedFlat, and diskANN indexes, VectorDistance queries with TOP N and filters, the change feed and its modes, the change feed processor, the Python pull model, and the Azure Functions trigger. Includes three original practice questions.
Preview the topics
- A container holds 300 product vectors, with a diskANN index. Searches cost many request units, and behave like full scans. Why?
- A worker keeps a search index in sync with a container, through the change feed in latest version mode. Products deleted from the container stay in the search index. What should the team do?
- A Python worker reads the change feed with query_items_change_feed. After a restart, it must continue where it stopped, not from the beginning. What should it do?