This repository is a collection of small AI agents that demonstrate various concepts and capabilities in agent-based AI. The projects in this repository evolve alongside my exploration and learning in the field of AI agents, and new agents will be added over time.
This agent helps generate and evaluate code implementations. It works as follows:
- Requests two different code implementations for a given task.
- Another language model evaluates both implementations and selects the better option, providing an explanation for the choice.
- This project demonstrates how AI agents can collaborate to improve code quality.
A Gradio-based chatbot that acts as your professional alter-ego, answering questions about your career, background, skills, and experience. Uses OpenAI's function calling to record user details and unknown questions.
This agent is designed to automate sending cold emails:
- Generates personalized emails for specified users using a language model.
- Evaluates the quality of the generated emails and improves them iteratively.
- Currently under development, with ongoing improvements planned.
Learning Assistant is a multi-agent system that automatically generates personalized educational content. The system uses four specialized agents working in sequence - from planning your learning path, through creating content and exercises, to publishing everything to Notion.
- Add more AI agents with different capabilities.
- Explore agent collaboration, evaluation, and autonomous decision-making.
- Continuously improve existing agents and integrate new AI concepts.
Each agent has its own folder with instructions on how to run and test it. Check the respective project folders for setup and usage details.
Contributions and suggestions are welcome! Feel free to open issues or submit pull requests to improve existing agents or add new ones.