What it does:

Powered by a retrieval-augmented generation (RAG) system, our system makes finding bylaws as easy as having a conversation, while remaining accurate, citing its sources, and clearly indicating when it doesn't have an answer.

How we built it:

uses gemini api with a vector database build with mongodb to get relavent queries for a given search.

Challenges we ran into:

  • Getting Relevant Results from our vector database
  • Setting Up the Database
  • Connecting the Database to Gemini

Accomplishments that we're proud of:

  • The effectiveness of the RAG system

What we learned:

  • How to extract data from PDF's
  • How to scrape websites

What's next for paralegal:

  • Find people in Toronto and collect user feedback
Share this project:

Updates