Sumanth077/Hands-On-AI-Engineering

A curated collection of practical AI projects implementing OCR systems, RAG, AI agents, and other AI use cases.

What is this?

Hands-On AI Engineering is a curated collection of practical, production-ready AI projects. It's a single repository where you can find working examples across many AI areas — from AI agents that automate tasks, to systems that read documents, understand audio, process images and video, answer questions using your own data (RAG), and even fine-tune models for specialized jobs.

Each project includes complete code, setup instructions, and documentation, so you can learn by doing and adapt the code for real-world use.

What's inside?

The README organizes projects into these categories:

AI Agents — The largest section. These are autonomous programs that perform tasks like researching topics, analyzing finances, booking travel, monitoring brands, reviewing code, and more. Many use multiple specialized agents working together.
OCR — Tools that extract structured data from images and documents, like receipts, handwritten prescriptions, or math formulas.
Audio — Projects for transcribing, translating, and understanding audio, including voice agents that answer phone calls.
Multimodal — Systems that combine vision, video, and language, such as summarizing YouTube videos, extracting text from images, or answering questions about charts.
RAG Applications — Retrieval-Augmented Generation systems that let you ask questions over your own documents, websites, or knowledge bases, with answers grounded in the source material.
Fine-Tuning — An example of training a model on a specific task (turning natural language into SQL queries for inventory databases).

Key features highlighted in the README

Learn by doing — every project has complete code and instructions.
Production-ready — projects follow best practices and are meant to be adaptable.
Diverse use cases — from simple OCR to multi-agent workflows.
Multiple model providers — uses OpenAI, Anthropic, Google, and open-source models.
Active community — regular updates and new projects are added.

How to contribute

The project welcomes contributions. To add a project:

Read the contributing guidelines (linked as CONTRIBUTING.md).
Create an issue proposing your project or improvement.
Put your project in the correct category folder.
Submit a pull request (one project per PR).

Each project must include a comprehensive README.md, a requirements.txt or pyproject.toml, and an .env.example for API keys. Folder names should use snake_case.

License

The repository is licensed under the MIT License (see the LICENSE file).

Who maintains it?

It's built by the AI Engineering Community (linked to a newsletter) and maintained by Sumanth (contact: sumanth@devable.ai).

All details above are taken directly from the README. No features or facts were added beyond what's stated there.