Unlocking Data with Generative AI and RAG

12

Combining RAG with the Power of AI Agents and LangGraph

One call to an large language model (LLM) can be powerful, but put your logic in a loop with a goal toward achieving a more sophisticated task and you can take your retrieval-augmented generation (RAG) development to a whole new level. That is the concept behind agents. The past year of development for LangChain has focused significant energy on improving support for agentic workflows, adding functionality that enables more precise control over agent behavior and capabilities. Part of this progress has been in the emergence of LangGraph, another relatively new part of LangChain. Together, agents and LangGraph pair well as a powerful approach to improving RAG applications.

In this chapter, ...