serverless-chat-langchainjs

Why we included this project

This is a practical blueprint for teams that want to turn internal documents into a conversational Q&A experience without rearchitecting their stack from scratch. It connects LangChain.js, Azure Functions, and Azure Static Web Apps so ingestion, vector search, and chat responses all run on a serverless backend, with Cosmos DB handling both session history and vector storage. The Contoso Real Estate sample data shows the whole pipeline working end to end before you swap in your own content, and the Ollama local-dev path lets you experiment without paying for cloud resources. Because it ships as a complete template with infrastructure-as-code and documented deployment steps, you adapt it rather than import it, which suits anyone building a production RAG chatbot who wants an opinionated reference to start from.