Inspiration
Every company has a lot of documents that are often scattered everywhere. In order to help workers find their way around documents more easily, reach the desired answer faster and more efficiently, we created CatAIog, which solves these problems. Also, CatAIog is not just for workers. It is also designed as a product catalog for a company so that their customers can find the desired product more easily.
## What it does CatAI is a local chat agent that answers questions which correspond to your internal documents that you submitted into docs folder.
How we built it
We used the Databricks Dolly-v2-3b model, in which we fed our internal knowledge about the documents using llama_index. The UI was built using the Streamlit library.
Challenges we ran into
Our first obstacle was the lack of an API for the open source LLM models. Another obstacle was allocating enough GPU RAM to run the models.
Accomplishments that we're proud of
We are proud to have successfully integrated the GPT index with the LLM model who is able to read multiple documents. And the fact that we made a simple user interface to make the user experience better.
## What we learned We learned what the GPT index is for and how it works. Also we learned how to integrate it with the LLM model.
Log in or sign up for Devpost to join the conversation.