We have all been blown away by ChatGPT. In fact, you’re probably wondering if I wrote this guide or if GPT wrote it for me.
Unfortunately, GPT has yet to distill my winning personality, so I've had to do this the old-school way.
Below, I’ll take you through how you can build your own conversational Q&A widget that is trained on your knowledge base.
By the end, you’ll get all of this for free:
- the frontend code for a ChatGPT style interface
- a notebook that ingests PDFs into a Redis database, with vector embeddings from OpenAI
- a Relevance AI chain template that powers the conversational Q&A
The beauty of it all is that these pieces can be easily tweaked to fit your specific needs.
Here's a sneak preview of the final product:
Updated 5 months ago
Here are the steps in this guide
- Scaffolding our frontend
- Hold up! What are “chains”?
- Setting up a dummy Q&A chain
- OK, so how will our actual Q&A chain work?
- Extracting your PDFs into Redis, with OpenAI embeddings
- Setting up the frontend for our Q&A chain
- Using chat history to power a conversation
- Adding references, like Bing Chat
- Adding some frontend sugar