How do I build a retrieval-augmented generation (RAG) pipeline on Databricks — what are the best pra

[Suheb]

How can I build an AI system (using Databricks) where the AI model doesn’t just rely on its built-in knowledge, but also retrieves real information from a database or documents before answering?

[KaushalVachhani]

@Suheb , We have a bunch of resources around Databricks RAG. Did you have a chance to look at the available documentation? Adding a few of them below.

Databricks docs for RAG: https://docs.databricks.com/aws/en/generative-ai/retrieval-augmented-generation

Databricks docs on Agents with tool calling: https://docs.databricks.com/aws/en/generative-ai/agent-framework/author-agent?language=LangGraph 

Dbdemos: https://notebooks.databricks.com/demos/ai-agent/index.html 

We also have MCPs and Agent bricks, which might be helpful. Feel free to search them as needed.