Secure Case Chat

A secure, AI-assisted case analysis platform where attorneys privately chat with an agent that can only access their own case data. The platform is built with security and auditability by default.

Comment

Inspiration

Attorneys are busy, often needing to quickly analyze an enormous amount of case data across many formats. We are using AI to ingest and analyze PDFs using Amazon textract, text documents using LLMs, and video using TwelveLabs video analysis MCP. The goal is to allow the attorneys to use natural language search to query across all formats to expedite their work in a secure way.

What it does

Unifying unstructured case data across formats into a single queryable interface. Code is secured by semgrep. We wanted to implement Vanta but didnt get to it.

How we built it

Tailwind, React, Next, Semgrep, and wanted to use AWS but the integration failed so we used vercel

Challenges we ran into

Integrating and making AWS side work.

Accomplishments that we're proud of

Finding a good use case, working as a team, and having fun despite challenges. Learning about semgrep. Tried working with AWS lambda functions and DynamoDB, but we pivoted well.

What we learned

Learned about semgrep and custom rules. Learned more about AWS. Learned that I need to shake off more coding rust.

What's next for Secure Case Chat

Going for venture capital funding... perhaps.

Built With

Share this project:

Updates