Cheapy - AI Hotel Finder
An AI-powered hotel search tool that compares prices across 100+ booking sites to find the best value stays. I built the backend and search infrastructure.
Overview
Cheapy aggregates hotel listings from over 100 booking platforms and uses AI to surface the best value options for a traveler’s budget, dates, and trip type. The idea is to replace the usual ritual of comparing prices across a dozen open tabs with one search, ranked results, and price alerts that actually fire.
What it does
- Searches 100+ booking platforms in a single query and ranks results by value, not just the lowest price
- Tailors recommendations to budget, location preferences, and trip type, whether business, leisure, or family
- Sends price alerts when a saved hotel drops in cost
- Publishes travel guides and booking tips to bring in organic search traffic
Stack
- Frontend: Next.js 14, React 18, TypeScript, Tailwind CSS
- Backend: .NET 8, ASP.NET Core, Entity Framework Core
- Search: Elasticsearch
- Caching: Redis
- Database: PostgreSQL
- Deployment: Docker, CI/CD with GitHub Actions
My role
I built the backend API layer and the search infrastructure. In practice that meant integrating third party hotel data sources, writing the price comparison and ranking logic, and tuning Elasticsearch queries to stay under a second of response time while handling real load. The Redis caching layer turned out to be the piece that made or broke the experience: without it, aggregating live data from dozens of sources was simply too slow to feel usable.