Code with Cosmos: Building AI Applications with Azure Cosmos DB
Marko Hotti and Patty Chow demonstrate how Azure Cosmos DB can be used as a single system for application data plus AI-oriented search, using a travel planner example.
Overview
The video shows how an AI travel planner can use Azure Cosmos DB to keep operational data (like inventory and booking information) alongside vector embeddings and search indexes, so the app can answer natural-language requests while still supporting transactional workflows.
Key capabilities highlighted:
- Storing application/operational data and vector embeddings in one database
- Full-text search for exact-match style queries
- Vector search for semantic similarity
- Hybrid search to balance semantic relevance with precision
- Transactional workflows so results reflect live inventory and booking data
Example scenario discussed:
- Finding a “quiet boutique hotel near Barcelona’s Gothic Quarter”
- Balancing semantic relevance (meaning) with exact-match constraints (precision)
- Using live operational data so recommendations align with availability and booking state
Links
- GitHub repo: https://aka.ms/cosmos-travel-planner
- Playlist: https://aka.ms/CodeWithCosmosPlaylist
- Try Azure Cosmos DB: https://aka.ms/trycosmosfree