Developed an end-to-end Retrieval-Augmented Generation (RAG) tutorial for Azure Cosmos DB in the langchain-ai/langchain-azure repository, focusing on reproducibility and ease of experimentation. The solution integrated Python, LangChain, and Gradio to demonstrate vector search, hybrid search, semantic caching, and per-user chat history within a self-contained Jupyter Notebook. A helper script enabled seamless sample data download, while configuration relied on local environment variables to avoid exposing secrets. All changes were isolated to the sample and tutorial workflow, maintaining library integrity. The work provided a practical, interactive resource for developers to evaluate RAG capabilities with minimal external dependencies or setup complexity.