DeepLearning.AI · 2024-05-07 · notable
Building Agentic RAG with LlamaIndex — Free DeepLearning.AI Short Course
Free 44-minute course by Jerry Liu (LlamaIndex CEO) teaching how to build RAG agents that reason over documents, use tools, and handle multi-document queries. Includes 4 runnable code labs.
A free, hands-on course that teaches you to build RAG agents that can reason, use tools, and query multiple documents.
Key specs
| Duration | 44 minutes |
|---|---|
| Lessons | 6 |
| Code labs | 4 |
What is it?
This is a free short course on DeepLearning.AI taught by Jerry Liu, co-founder and CEO of LlamaIndex. In 44 minutes across 6 lessons, it walks you through building agents that go beyond simple retrieve-and-generate. You build router agents that decide whether to do Q&A or summarization, implement tool calling where the LLM infers function arguments, create multi-step reasoning loops, and extend agents to work across multiple documents.
How does it work?
The course uses LlamaIndex's agent framework. You start with a basic RAG pipeline over arXiv papers, then progressively add agentic capabilities: routing queries to the right tool, chaining multiple retrieval steps, and having the agent decide when it has enough information to answer. Each concept comes with a runnable code example in the browser.
Why does it matter?
Agentic RAG is where most production RAG systems are heading — simple retrieve-and-generate pipelines hit a ceiling when queries need reasoning across sources. This course is one of the fastest ways to understand the pattern, and it is free. Jerry Liu teaching it means you learn the framework from the person who designed it.
Who is it for?
Developers building RAG systems who want to add agent capabilities.
Try it
https://www.deeplearning.ai/short-courses/building-agentic-rag-with-llamaindex/