Azure Cosmos DB TV - From Burger to Bots: Agentic Apps with Cosmos DB and LangChain.js | Ep. 111

On this episode of Azure Cosmos DB TV, host Mark Brown is joined by Yohan Lasorsa to explore how you can build modern, agent-powered applications on a fully serverless foundation. This session looks at how a single data platform can support both transactional workloads and AI-driven experiences without adding extra infrastructure or operational complexity. You’ll see how application data, APIs, and AI agents come together in a practical, end-to-end example you can adapt for your own projects. See how Cosmos DB powers both transactional features and AI-driven experiences without extra infrastructure. We’ll walk through a practical sample: a REST business API for orders, an MCP bridge that lets an agent access business data, and LangChain.js for the agent with chat memory. All on a serverless footprint you can ship quickly! What we’ll cover in this session: • Introduce the session with the use case, present the sample and its architecture, and explain where Azure Cosmos DB fits into the overall design. • Demo of the sample application, followed by a walkthrough of key sections of the codebase. • Discussion of possible extensions to this use case, including relevant Azure Cosmos DB features you can apply as you scale. Learn more and explore the resources used in this session: https://aka.ms/lcjs/agent-mcp Useful links: • Azure Cosmos DB TV - https://aka.ms/AzureCosmosDBLiveTV • Subscribe to Azure Cosmos DB on YouTube -    / azurecosmosdb   • Follow Azure Cosmos DB on X -   / azurecosmosdb   • Follow Azure Cosmos DB on LinkedIn -   / azure-cosmos-db