Writing
What actually happens when AI agents and automated workflows go in front of people whose money is on the line. Mostly the parts that don’t make it into the launch post.
Everyone’s instinct is to hand the whole matching problem to the model. We gave it the messy half and kept the ranking deterministic — here’s the argument, and the part we still haven’t solved.
An agent in production failed in a way no eval suite would have caught — because the thing that changed wasn’t the model, and the thing that broke never threw an error.
Work with me
I take on a small number of outside engagements — production AI agents and workflow automation, the parts that have to hold up when a real operator is depending on them.
Fixed-scope sprint
4–8 weeksAn agent or automated workflow shipped to production, with the tool-boundary validation, contract tests, and approval gates that make it trustworthy.
Seed / Series-A teams with a workflow that should have been automated last quarter.
Fractional technical lead
Ongoing, part-timeArchitecture, agent reliability, and technical direction without a full-time hire. Design calls, reviews, and hands-on work where it counts.
Teams shipping AI features who need someone senior on it continuously.