I  

I’m a dev who spends most of his time somewhere between AI systems, LLMs, and software infrastructure. I’m particularly interested in understanding what happens underneath the abstractions - from model inference and memory to the systems that make AI actually useful.

Most of my work revolves around LLM applications and agentic systems: local-first AI, model serving, RAG, MCP, tool use, orchestration, and building agents that can interact with real systems instead of just talking about them.

Lately, I’ve been going deeper into AI systems engineering - running and optimizing models locally, experimenting with quantization and inference, and figuring out how models, tools, memory, and infrastructure fit together end to end.

I also spend time building in Web3, developer tooling, and whatever strange idea I can’t stop thinking about. A lot of my projects start as experiments and end up becoming actual products.

I have a habit of going down technical rabbit holes just to understand how something works, then trying to build a smaller version myself.

I like building things that shouldn’t work, making them work, and then figuring out why.