Better AI Responses with RAG Fusion

This video is about getting potentially better results from your LLM-driven applications using a method called RAG Fusion.

LINKS:

Code for Implementing RRF in node.js (Gist): gist.github.com/heaversm/4fefbf62b01ac3e0eb35d31a3…

Original Article on RAG Fusion: towardsdatascience.com/forget-rag-the-future-is-ra…
Colab Notebook - Test RAG Fusion: colab.research.google.com/drive/1VaENoPcSznXRjiOzb…
Reciprocal Rank Fusion research paper: plg.uwaterloo.ca/~gvcormac/cormacksigir09-rrf.pdf
Original RRF Code Sample: github.com/Raudaschl/rag-fusion/blob/master/main.p…
Pinecone (for Colab API Keys): app.pinecone.io/
Open AI (for Colab OpenAI API Keys): platform.openai.com/overview

About Me: I am a Staff Design Technologist on Mozilla's Innovation Team. All opinions and explorations are my own. Learn more about Mozilla Innovation at future.mozilla.org

0:00 intro to RAG fusion
1:04 research paper on reciprocal rank fusion
1:30 RAG Fusion process
2:10 RAG Fusion applied to an LLM Query in Podquest
2:50 Code output during generation
3:20 Comparing results
4:35 Google Colab Jupyter notebook with Python / Langchain demo
6:35 RAG Fusion adoption