Inspiration

Replying to messages can feel overwhelming when life gets busy. We wanted to solve that.

What it does

Digital Twin reads your past texts and replies like you would — instantly.

How we built it

We used Python, Streamlit, LangChain, and Gemini. Built in Cursor. Just the two of us (Uvernes Somarriba and Abhi Chauhan).

Challenges we ran into

Time. Also balancing generation quality with speed.

Accomplishments that we're proud of

It actually sounds like us. And it works end-to-end.

What we learned

Prompting is an art. Streamlining AI into simple UI takes work.

What's next for Digital Twin

Adding memory, better context handling, and mobile support.

Built With

Share this project:

Updates