About Catchup

What is Catchup?

Catchup scrapes 15 of the top tech engineering blogs, extracts their articles, and converts them into NotebookLM-style conversational podcasts featuring two AI hosts. The result is a Spotify-like web player where you can browse, filter, and listen to the latest thinking from companies like Meta, Uber, Airbnb, Cloudflare, Netflix, and more.

Why?

Engineering blogs are some of the best sources of real-world technical knowledge, but keeping up with all of them is time-consuming. Catchup turns long-form posts into bite-sized audio you can listen to during a commute, workout, or coffee break — so you never fall behind on what top engineering teams are building.

How it works

1. Discover & Crawl — Sitemaps, RSS feeds, and blog page scraping find new posts across all sources.
2. Extract & Tag — Content is extracted with quality scoring, then auto-tagged into 12 categories (infrastructure, AI/ML, security, and more).
3. Generate Podcasts — An LLM writes a two-host conversational script, then text-to-speech produces the final audio.
4. Listen — Browse the web player, filter by source or tag, build playlists, and hit play.

Tech highlights

• Crawl4AI + RSS + sitemap discovery for broad coverage
• Multi-strategy content extraction with quality gating
• OpenAI TTS for natural-sounding conversational audio
• FastAPI backend with async job tracking
• Next.js 16 + React 19 frontend with global audio player
• Deployable via Docker or Railway with persistent storage

Built by

Catchup is a personal project by Gaurav Surtani, built to make engineering knowledge more accessible. Contributions and feedback are welcome on GitHub.