🧠 I create evolving, goal-driven AI systems that orchestrate tools and agents to perform complex tasks — autonomously.
I'm the creator of the Evolving Agents Toolkit (EAT) — an open-source Python framework for building adaptive, multi-agent systems designed for AI-First workflows.
- EAT: Build systems that go beyond prompt chains. With a central SystemAgent, SmartLibrary, SmartAgentBus, and evolution tools, EAT enables true agent orchestration and self-improvement.
- Designing architectures that learn, adapt, and improve based on feedback — with modular integration of OpenAI, BeeAI, LangChain, and more.
- Exploring use cases in automation, project management, educational content creation, and complex reasoning workflows.
- Created Project Copilot, an AI Project Manager that turns conversations and documents into structured backlogs.
- Built 3D medical visualization systems, ML models for healthcare, and prompt workflow engines before it was cool.
I'm always up to discuss AI agents, adaptive systems, or open-source collaboration.
Find me on Twitter: @matiasmolinas