I'm an AI-native builder exploring how AI can move beyond answering questions and become a reliable collaborator — one that understands context, uses tools, and helps turn intent into action.

I'm interested in the space between a promising model demo and a system people can trust every day. That is where agents need better skills, memory, orchestration, and judgment.

Most of my work lives in private repos — happy to share on request.

• 🧩 AI agents & skills — Composable systems that can reason, act, and complete useful work instead of stopping at an answer.
• 🧠 Personal AI — Assistants that grow more useful through context and continuity while keeping people in control.
• 🛠️ AI-native tools — Practical infrastructure for building, governing, and understanding agent systems without hiding the hard parts.

The future of AI is not just a bigger chat box.

I care less about impressive demos than systems that remain useful after the novelty wears off. Good AI should be understandable enough to guide, dependable enough to delegate to, and flexible enough to evolve with the people using it.

The goal is not to automate everything. It is to give human attention back to the things that deserve it.