Q&A Bot Agent

I built an Elasticsearch Q&A Bot for the Agent Builder Hackathon. It uses vector search to turn PDFs into instant, context-aware answers with multi-turn chat and MCQ generation.

Comment

Inspiration

There is lot of knowledgeable information exists in an archive, which can be represent as a nice Q&A form, so that when user complete the answer It will give correct answer with some stories behind it.

What it does

Collect the information and transform that information into Q&A using LLM, and build a nice short story to give awesome experience to user.

How we built it

Using React UI, Python Backend, Ollama LLM to generate Q&A

Challenges we ran into

To build using everything localhost and opensourced.

Accomplishments that we're proud of

We have given a nice representation of the knowledge into a great learning experience.

What we learned

Learned about various LLMs available in the market and their pros and cons.

What's next for Q&A Bot

This Bot can be fructify using various archival, which is already present there just need good interesting stories to share across user friendly opinion.

Built With

Share this project:

Updates