Architecting AI Workloads with Cosmos DB: Azure vs Fabric

Cosmos DB is increasingly used in AI applications and is now available in Microsoft Fabric. But when should you use it—and which version? Whether you’re implementing simple RAG patterns or orchestrating complex multi-agent workflows, Cosmos DB can play key roles: as a vector store, for persisting conversation history and agent state, and as a serving layer for real-time inference.

In this session for solution architects and developers, we’ll explore:
• How Cosmos DB supports AI workloads through practical architecture patterns.
• When to choose Azure Cosmos DB versus Cosmos DB in Microsoft Fabric, including key differences and decision criteria.

You’ll leave with a clear understanding of how Cosmos DB fits into AI applications and how to select the right service for your scenario.

• Understand Cosmos DB’s roles in AI — Learn how it supports vector storage, conversation history, agent state, and real-time inference in AI workflows. • Compare Azure vs Fabric Cosmos DB — Discover practical differences, performance considerations, and integration scenarios for each service. • Make informed architecture decisions — Gain clear criteria for selecting the right Cosmos DB option for your AI application, from RAG patterns to multi-agent systems.