AI · GitHub repo
ai-engineering-hub
Open-source tutorials and project templates for LLMs, RAG, and AI agents.
ai-engineering-hub is an open-source GitHub repository containing over 90 practical AI engineering projects organized by difficulty level. It provides complete Jupyter notebooks and codebases covering OCR, RAG pipelines, model evaluations, fine-tuning, and multi-agent systems using frameworks like CrewAI and the Model Context Protocol (MCP). It is a direct reference collection for developers looking to build and deploy practical AI applications without starting from scratch.
What it's for
- Build custom RAG pipelines over PDFs, audio, codebases, or databases
- Develop multi-agent research and automation workflows using CrewAI and MCP
- Deploy local chat and reasoning interfaces with open-source LLMs
- Fine-tune models like DeepSeek using Unsloth and Ollama
Features
- 90+ ready-to-run projects categorized into Beginner, Intermediate, and Advanced tiers
- Implementations for basic and agentic RAG with various vector stores and embeddings
- Extensive Model Context Protocol (MCP) integrations for tools, memory, and search
- Vision and OCR setups using Llama 3.2 Vision, Gemma-3, and Qwen 2.5 VL
- Multi-agent automation workflows for content creation, research, and analysis
- Model evaluation and observability examples using CometML Opik
- Includes an end-to-end AI Engineering Roadmap for structured learning
Pricing
Open source. Free (MIT License)
Details
- Type
- GitHub repo
- Category
- AI
- Topic
- AI Development
- Open source
- Yes
- Platforms
- Linux, macOS, Windows
- For
- AI engineers, data scientists, and developers building LLM and agent applications.
- Tags
- llms, rag, ai agents, mcp, fine tuning, crewai, python
Why it was saved
- @nathanhodgson.ai: It provides full working source code and Jupyter notebooks under an MIT license so developers can build production-ready AI applications without starting from scratch.
- @akshay_pachaar: A working reference for combining a web-scraping infra (Bright Data) with multi-agent orchestration (CrewAI) and a quick UI (Streamlit), and the demo shows exactly what the end product looks like.
- @akshay_pachaar: Shows the component breakdown and data flow of the app in one picture; a searchable pattern for building scraper + agent pipelines.
Related
Design inspiration MCP server and archive for AI coding agents.
Open-source capability layer giving AI agents access to browse and scrape platforms
Autonomous harness engineering framework that optimizes AI agent setups overnight
Curated collection of skills, tools, and plugins for Claude workflows
Step-by-step tutorial wiring Hermes Agent to Box for cited enterprise retrieval
Open-source multi-agent framework for building autonomous social and on-chain AI agents
Updated