Why I work this way
Operational experience came before AI.
Before LaunchSoloAI, I spent five years leading high-stakes public-sector operations involving multidisciplinary teams, long campaigns, crisis communication, and procurement under public scrutiny. That work taught me that a process is only as strong as its ownership, evidence, handoffs, and behavior under pressure.
I bring the same discipline to AI systems. A fluent answer is not an operating model. A successful demo is not a controlled deployment. Every engagement begins with the real workflow, the people accountable for it, the failure modes, and the evidence required to trust the new path.
What I build
Workflow automation, tool-using agents, retrieval systems, model routing, evaluation and monitoring, agent cost controls, AI product stabilization, and the integrations around them. Runcap is my public open-source work on cost and proof boundaries for AI coding agents.
How engagements work
Remote and async from Calgary, available worldwide. The first deliverable is written: the current bottleneck, automation boundary, human decisions, acceptance test, open risks, and a fixed scope only when implementation is justified.