Jarvis AI

Inspiration

We wanted a lightweight personal AI assistant that runs in the browser, is customizable, and doesn’t depend on expensive cloud services — something anyone could build, run, and improve.

What it does

JARVIS AI Assistant processes natural language commands, automates tasks, and interacts with users through a smart frontend and a powerful Python backend. It's fast, offline-friendly, and extendable.

How we built it

We used React, TypeScript, and TailwindCSS for the frontend, managed by Vite. The AI logic is powered by Python, with Bolt.new used to scaffold and organize the project quickly and cleanly.

Challenges we ran into

*Integrating Python with the frontend in a browser-hosted setup *Designing for both speed and simplicity *Ensuring cross-platform deployability without relying on cloud APIs

Accomplishments that we're proud of

*Fully functional AI interface without cloud dependence *A clean, modern UI powered by Vite and Tailwind *Quick setup and deployment using Bolt.new scaffolding

What we learned

*How to balance frontend and backend separation in hybrid projects *Better ways to design with performance and offline usage in mind *The power of fast dev tools like Bolt and Vite

What's next for JARVIS AI Assistant

*Voice command integration using Vosk *Hosting the Python backend with ngrok or Railway *Packaging it into a local desktop or mobile app *Adding memory and plugin support