// the find
ishandutta2007/Top-AI-repos
🎖️🏆🏅 193 most popular open source AI projects 🌟 Star if you like it! 🌟
A hand-maintained directory of ~193 popular AI/ML repos, split into eight categories (agents, RAG, inference serving, etc.) and sorted by star count, with a GitHub Action that keeps the ranking fresh. It's for someone who wants a quick lay-of-the-land snapshot of what's popular in AI tooling right now, not for anyone looking to depend on code from it.
The category split (agents-and-automation.md, rag-and-data.md, etc.) is more useful than a single flat table — you can actually scan by problem domain instead of scrolling 193 rows. The auto-sort.yml workflow re-pulls live star counts so the ranking doesn't go stale the way most awesome-lists do after the first PR. Columns for creation year and primary language give context a plain star count doesn't.
It ranks by star count, which rewards marketing and timing over merit — several entries near the top are recent viral repos, not necessarily the best tool for the job. Every description is a one-line lift from the GitHub repo's own tagline, so there's zero independent judgment on what's actually good or where each project falls short. The README is mostly badge markup (four shields per row, custom star-counter images) which makes it slow to load and noisy to read on GitHub itself. The maintenance scripts (dedup.py, fix_serials.py, sort_table.py) are ad hoc one-offs with a single smoke test, so contributing a fix means understanding an untested pipeline rather than a documented process.