Enterprise-grade AI chatbot architecture on Azure integrating Claude LLM via MCP Server, combining Azure Data Factory ETL pipelines, Event Hubs streaming, Cosmos DB knowledge store, and Azure AI Search for intelligent retrieval. Data flows from multiple sources through ingestion and processing subnets into Synapse Analytics and a knowledge store, enabling the MCP Server to orchestrate Claude API calls with cached context and real-time search results. This pattern demonstrates secure, scalable AI application design with Key Vault secrets management, API Management gateway protection, and WAF policies safeguarding the chatbot frontend. Fork this diagram to customize subnet ranges, add additional LLM providers, or adjust data pipeline stages for your enterprise knowledge base requirements.