Physics-Grounded Founders.
Physics-Grounded AI.

Three decades building real operational systems in industrial automation, computer vision, process control, and systems integration on the steel floor. The kind of experience that comes from making things work in the melt shop and the crane bay — not from benchmarks.

The core question driving this work: how do you ground AI reasoning in plant physics — BOF kinetics, heat transfer, caster solidification, crane kinematics — so it can be trusted on the floor? The answer is architectural. GoatAI is the answer.

Physics-grounded agentic reasoning is not a research hypothesis. It is what you build when you've spent 30+ years watching software miss the physics of the plant it's embedded in.

AI researcher and systems developer focused on the gap between theoretical AI capability and what actually works in physical environments. The challenge isn't building models — it's building reasoning architectures that respect plant physics, equipment geometry, and real-world operating constraints simultaneously.

Current focus: physics-AI integration research on steel-process physics — BOF kinetics, heat transfer, caster solidification reduced-order models — and building the agentic orchestration layer that makes multi-agent reasoning over the plant operationally viable.

The question isn't whether agents can reason about a heat or a cast. It's whether the reasoning is grounded enough in plant physics to be trusted when it matters on the floor.