An AI search engine over 50,000+ tutorials
A full-stack, AI-powered product that indexes and semantically searches more than 50,000 developer tutorials, built solo from zero.
- Role
- Creator & Full-Stack Developer
- Year
- 2023
- Hosting
- Cloudflare Workers and D1, with a Python indexing pipeline
- 50,000+
- Tutorials indexed
- AI
- Semantic search
- 100%
- Built solo
Overview
Developers waste time comparing tutorials scattered across platforms, with no single place to weigh ratings, price, and level. TutorialSearch aggregates more than 50,000 tutorials and lets you search across them by intent, not just keywords.
I built the entire product alone: the Python pipeline that scrapes and indexes tutorials, the AI-powered search and ranking, the React front end, and the Cloudflare deployment. It is a working example of an AI product taken from idea to launch.
What the build involved
- Built a Python data pipeline to aggregate and index 50,000+ tutorials from across the web.
- Implemented AI-powered semantic search with filtering by skill level, duration, and rating across 45+ categories.
- Shipped the React and Next.js front end and deployed everything on Cloudflare Workers with a D1 database.
Challenges & solutions
Challenge
AI search over tens of thousands of items had to stay fast and cheap.
Solution
Pre-computed embeddings, edge caching, and D1 at the edge so queries return quickly without a large inference bill.
Challenge
Tutorial data was messy and scattered across many sources.
Solution
A resilient Python pipeline that normalises, deduplicates, and categorises listings into one clean, searchable index.
Available for new work
Want results like these?
Tell me what you are building. I will reply with a clear scope and a fixed quote.