In the last module you wired up a RAG pipeline step by step. In practice most teams use a framework for that glue code. The two most popular are LangChain and LlamaIndex, and they have different strengths.
LangChain is a broad toolkit for building LLM applications. It gives you prompt templates, chains that connect several steps, agents that can call tools, memory, and document loaders. LlamaIndex focuses on your data: loading documents, building an index, and answering questions through a query engine, with sensible RAG defaults built in. The two can also be used together.
The first two videos introduce LangChain, and the third shows LlamaIndex building an agentic RAG app. Two of the LangChain videos date from 2023, so some import paths and class names have changed since. Focus on the concepts and check the current documentation when you write your own code.
A beginner walkthrough of LangChain that builds a restaurant idea generator with Streamlit, and touches on more advanced features such as retrieval question answering.
Video 46minA quick tour of the core concepts of the LangChain framework, with the code available as a notebook so you can follow along.
Video 20minA code and workflow walkthrough of LlamaIndex that builds an agentic RAG app, where an agent decides how to search your data to answer a question.
Video 40min