Hi, I'm Brad

I build AI agent teams that replace a roomful of experts: the research, analysis, and judgment calls that usually take a group of experienced specialists.

I start with the people doing the work and go through real cases with them. I map the evidence they weigh, how they reach a decision, and what they do when a case doesn’t fit the usual pattern. Then I design and build the system: the models, agents, integrations, and evaluation, along with the points where an expert should review the result.

I’ve been working in applied AI for over a decade. I trained my first neural network on receipt processing at Sensibill in 2015. I built Kwola, an open-source system that used reinforcement learning to explore web applications and find bugs, and at Checksum.ai I built models that generate tests from real user sessions.

I’m interested in Staff and Principal Applied AI roles where I own a hard problem from discovery through a system the team relies on.