Software builder, business angel, and writer working on AI tools and agent-operated knowledge systems.
I am building Commonplace, a self-hosting LLM wiki for turning AI-assisted work into durable, inspectable, linked knowledge.
The broader question I am exploring is whether AI-assisted thinking can accumulate: not just chat transcripts, but reusable claims, sources, procedures, reviews, and open loops that other people and agents can inspect.
The site is structured as a public mind interface:
- Entry point - current map of the work
- Cup - what I am looking for now
- External Minds - communities and projects connected to this work
- Archive - older ML Adventures writing
- AI Adventures - my Substack on AI, software, and agentic systems
- Commonplace - self-hosting LLM wiki
- llm-do - on hold while I focus on Commonplace first



