Agentic RAG Using CrewAI & LangChain!
Vložit
- čas přidán 11. 09. 2024
- Retrieval-Augmented Generation (RAG) is the process of optimizing the output of a large language model using an external knowledge base attached through a vector database.
It is okay until we have simple documents and use cases.
But, with the introduction of AI agents, the researchers thought why not integrate these AI agents in our RAG pipeline.
Because these AI agents are able to plan, come up with logic, reasoning, and they also know how to use tools whenever there is a requirement.
This gave rise to so called 'Agentic RAG'
AI Agent RAG, or Agentic RAG, utilizes intelligent agents and advanced LLMs to streamline and enhance the retrieval and generation process, making it more efficient compared to traditional RAG approach.
Let's build a simple agentic RAG workflow using LangChain, CrewAI and Groq. The complete notebook code is here: github.com/pav...
The video is inspired by 'AI Anytime': • Build an Agentic RAG u...
Nice video, it would be helpful to paus a bit on the notebook when u r explaining the agent and tasks - go through the description and expected outcomes a bit slowly.
Thanks for the feedback. I will really keep this in mind while I do my next set of videos. Thanks :)
Nice video 👍
Thanks 🙏
Pavan, question why do we need CrewAI here ?
CrewAI is the one that is helping us build these agents. It is the automation framework that combines these AI Agent. If you see the code, we are defining the flow with rag_crew to basically automate these ai agents.
@@pavanbelagatti any alternative agent please list out them
@@vamshiprince969 You have AutoGen, Agency Swarm and LangGraph
@@vamshiprince969if you are looking for any alternative tech/framework of crewAI then you can checkout Autogen
@@vamshiprince969 crewai is not agent. Its a framework, you can try autogen, langchain