Confidential - AI-Powered Review Aggregation Platform
A platform that unifies business reviews from multiple sources into one trusted score using AI pipelines. Built to replace slow manual work and inconsistent cross-platform ratings with automated, intelligent aggregation.
The problem
Reviews were fragmented across platforms with inconsistent formats, wrong labels, messy multilingual data, and difficult brand matching - making consumer choices slow and ratings unreliable.
What I built
Built a FastAPI backend on AWS with PostgreSQL, implemented AI pipelines using GPT-4o and NLTK for review summarisation and brand classification, and developed entity resolution logic to accurately match the same business across review platforms.
Outcome
- Unified reviews from Google, Yelp, and Trustpilot into one trusted score
- AI-powered brand matching and entity resolution across platforms
- Replaced manual spreadsheet work with automated classification pipelines
- Scalable AWS architecture handling large data loads with queues and caching
Stack
Running into something similar?
Start with a fixed-scope audit of your Supabase/Postgres app. If there is nothing worth fixing, I will tell you that too.