AI Engineering Hub
A collection of 90+ hands-on AI projects and tutorials covering LLM apps, RAG, AI agents, MCP servers and fine-tuning, mostly in Python.
AI Engineering Hub (~41k stars, no tagged releases, updated Nov 2026) is a GitHub collection of 93+ example projects sorted into beginner, intermediate and advanced, from local OCR and chat apps to RAG pipelines, multi-agent systems, MCP servers and fine-tuning notebooks. The code is MIT. Each project is set up separately with Python; some run fully on your computer with Ollama models, while many need API keys for hosted services such as OpenAI, Gemini, AssemblyAI or Firecrawl. It is learning material, not one app, and the READMEs promote the authors' newsletter.
The facts
| Question | Answer | Source | Checked |
|---|---|---|---|
| Runs on my machine? | Partly. Varies by project; some run fully locally with Ollama models, many call hosted AI APIs | github.com | Checked today |
| Code licence | Open source. MIT | github.com | Checked today |
| Model weights licence | With conditions. No weights in the repo; projects pull third-party models (Llama, Gemma, DeepSeek, Qwen and others) under their own licences, or call hosted APIs | github.com | Checked today |
| Commercial use | With conditions. The code is MIT, so commercial use is allowed; the models and hosted services each project uses have their own terms | github.com | Checked today |
| Paid services needed? | Optional. Varies by project; local Ollama projects need none, but many need API keys for hosted services such as OpenAI, Gemini, AssemblyAI, Firecrawl, Zep, Bright Data or Linkup | github.com | Checked today |
| Hardware | Unknown. Not documented (checked the main README and the llama-ocr, simple-rag-workflow, video-rag-gemini and notebook-lm-clone READMEs); varies by project and by model size | github.com | Checked today |
| Operating systems | Python projects; many READMEs give Windows and macOS/Linux setup steps, but not every project covers every system | github.com | Checked today |
| Setup effort | Medium effort, done per project; install Python packages, often Ollama or API keys, then run a Streamlit app or notebook | github.com | Checked today |
Each fact has its own date. Unknown means nobody has published it. We read their documentation; we didn’t install it. Running it yourself costs your hardware, electricity and setup time. Report a correction.
No licence changes since we started watching. Licence changes: follow AI Engineering Hub and we email you if its terms change.