I design and build production agentic workflows, multi-agent evaluation loops, and deterministic controls around probabilistic LLMs. I spend my time executing in the real mess of agent harnesses, tool orchestration, and automated test pipelines.
Professionally, I bring more than 15 years of enterprise-level software engineering and architectural leadership across multi-disciplinary, distributed groups shipping everything from core web applications to high-stakes marine survey and asset-integrity systems.
My focus at front left is closing the gap between AI hype and production reality: designing the context, automated harnesses, and verification boundaries required to make AI agents reliably do useful work.
