LLM researcher focused on reinforcement learning, agent systems, and practical AI infrastructure.

I build systems for agents that remember, coordinate, use tools, and operate in real workflows. I care about turning AI agents from demos into reliable products people can use every day.

• AI / Agents: LLM agents, tool use, memory systems, MCP
• Languages: TypeScript, Python
• Product: React, human-agent interfaces, automation workflows
• Research: Reinforcement learning, evaluation, long-context systems

• Reinforcement learning for LLM agents
• Long-running memory and context systems
• Reliable tool use and workflow automation
• Personal AI interfaces

Check my pinned repositories below to explore agent systems, personal AI interfaces, and automation tools.