- P 3/19 Muhammadia Colony Susan Road
- info@aisolutions.pk
-
Opening Hour Mon - Sat 9:00 - 6:30, Sunday - CLOSED
Book Recommendation System
Need Tech Service?
Contact Us
Professionally reintermediate technics Credibly pontificate turnkey
Book Recommendation System
A full-stack web application that delivers personalized book recommendations using machine learning. It combines a React frontend, a FastAPI backend, and a KNN-based content filtering model, with Redis caching layered in to keep API responses fast.
The system uses TF-IDF vectorization to compare book titles, authors, genres, and descriptions, then applies K-Nearest Neighbors to surface similar titles from a catalog of roughly 10,000 books. Redis caching cuts response times from 50-100ms down to 5-10ms, reduces database load by 50-80%, and pre-warms popular books on startup, while still degrading gracefully if Redis is unavailable.
Users get secure authentication via Supabase, book search and discovery, saved favorites, and personalized suggestions, all through a responsive Tailwind CSS interface. An admin panel gives librarians full CRUD control over the book catalog. The backend is deployed on Render with a PostgreSQL database on Supabase, and the frontend is deployed on Vercel with live interactive API docs available.