Building continually learning systems for AI-native teams.
I’m an AI founder building products solo. I’m interested in data flywheels, human-in-the-loop design, and agents that actually improve through use. I write about what I’m learning as I build.
In past I have worked on AI for satellite imaging and tools to produce data for training and evaluating LLMs.
Projects
Some things I've worked on:
- ComposeKit — AI-native distribution pipeline for builders.
- Spawnbase — Turn recurring work into reliable AI workflows you describe in plain language.
-
Skibble — A daily digest that pulls signal from Reddit, X, and Hacker News into one read. - Knowlio — AI exam prep that turns your course materials into custom practice tests.
Featured
-
Turning Agent Feedback Into Persistent Context
Making agents more effective isn't about chasing the latest model. It's about systematizing the feedback you're already giving and turning session learnings into persistent context.
-
Trends shaping technology in 2026
Where software value moves when agents eat the rest.
-
MCP Changed How We Think About Integrations
Early pain points, rapid progress, and why CLI tools don't make MCP obsolete.
Recent Posts
-
The Agent is the Runtime
A new computing paradigm is emerging where the agent is the runtime, natural language is the interface, and skills are starting to look a lot like apps.
-
How to Monitor Your LLM Conversations
Without evals, you're mostly guessing whether your AI product is working. Here's how to build a monitoring system that actually tells you.