Inspiration
My inspiration was coming from job posting. I've seen some legal firms searching for Agentic AI builders to enhance their securities and privacies with their clients, and I thought: Why not building an agent where lawyers could prepare clients' legal cases and communicate with them inside ONE system?
What it does
It's job is having a chatbot UI with multiple LLMs as selections for lawyers to prepare legal cases based on their preferences, generate their discussion with LLMs as PDF and send it to their clients.
How we built it
Crafted with React and Typescript, with Tailwind CSS polishing the UI
Challenges we ran into
Probably be the FastAPI Implementation. At the beginning, I tried starting with my backend. I wasn't able to see any log when I deploy on the docker. I had to shift my focus on building frontend to receive some feedbacks.
Accomplishments that we're proud of
I was really proud to see my Authentification goes through external application, authorize it and redirect BACK to my page.
What we learned
If I'm planning to build a full-stack application in future Hackathon. My roadmap should be: Brainstrom ideas -> Generate front end UI -> Discussion Backend Scale -> Crafting Frontend -> Testing functionality -> Crafting Backend -> Testing functionality -> Deploy on Docker
I missed some sections of roadmap and it caused me more time going back to work on it.
What's next for LegalAI
If possible, I want to build it as full-stack application with FastAPI as backend AND Docker deployable.
Built With
- auth0
- jspdf
- react
- typescript
- zustand
Log in or sign up for Devpost to join the conversation.