Obsessed with building systems that ship valuable output.

9+ years across two throughlines: retrieval/AI infrastructure and founder-operator work. Started in 2017 funding a NLP geopolitics research platform, Damocles Analytics, which received a Google News Initiative grant. It did document Q&A over news corpora — Haystack, FAISS, ElasticSearch, Neo4j — years before "RAG" was a household term. That work got me a Financial Times data science consultancy building news classification and ESG-ranking ML APIs.

Since then: senior engineering across fintech (payment settlement engines processing weekly merchant chargebacks) and several 0→1 builds, then founding engineer at Morpheus (mor.org), a decentralized AI inference marketplace on Base L2, where I built the API gateway and the full marketplace frontend from scratch, plus a wallet-first gateway (agentaigateway.com) that lets autonomous agents pay for their own inference, no human in the loop.

Current build: Nuclear News Network, an AI newsroom I run largely through agent pipelines. In two months it hit 11K AI-assistant citation events and page-1 Google and Bing rankings against incumbents with a decade of domain authority. It's less "media project" and more proof that agentic content systems can produce output clean and well-sourced enough to get cited by other AI systems at scale, which is the same property that matters for enterprise RAG.

I also ship in public: 25 published agent skills (1,100+ installs), open-source tooling with 500+ GitHub stars, upstream contributions to OpenCode and other projects, and an MCP server or two whenever a workflow gets annoying enough to fix once and reuse.
Through my independent practice I take on architecture assessments and infrastructure builds — inference deployment, RAG systems, agent orchestration, edge/disconnected AI — for teams who want something running, not a strategy memo.

If you're evaluating agentic AI, local/edge inference, or retrieval architecture and want someone who's shipped end-to-end, that's what I do.