ReVibe
- Event
- Google Ireland Hackathon by BGN
- When
- Oct 2025
- Result
- Selected from 500+ applicants
- Python
- FastAPI
- Gemini API
- Cloud Run
- PostgreSQL
Thrift stores are full of great clothes that nobody can find, and most people never think to look. ReVibe connects shoppers with local thrift stores and makes second-hand the easy option, built at the Google Ireland Hackathon by BGN after being selected from more than 500 applicants.
What we built
- AI outfit discovery: upload a photo or describe a look, and Gemini generates outfit ideas.
- Photo-based inventory upload and a digital inventory dashboard, so stores can get their stock online without the admin.
- Sell and donate flows for pre-owned clothing, plus a proposed "cash-for-clothes" model to grow supply and strengthen the circular economy.
My role
I built the backend: REST APIs in FastAPI for image and text prompt uploads, and the outfit generation service with Gemini 2.5, wiring image-to-text and text-to-image pipelines together, running on Cloud Run with PostgreSQL.
Built with Hari Mohan, Thohiru Omoloye, Hoang Le Lai and Bhumit Kaushik, with business guidance from Izuchukwu Echesirim.