ABOUT / OPERATOR · ARCHITECT · BUILDER
I work where AI strategy meets execution.
Enterprise AI gets interesting after the prototype. My focus is the system around the model: which problems deserve AI, how capabilities are architected, how teams govern and evaluate them, and whether the economics survive production reality.
01 / HOW I WORK
Equally comfortable with executives, solution teams, and builders.
I’m most useful in the space between business transformation and technical delivery. I can work with executives to define the problem, with solution teams to shape the approach, and with delivery teams to make it real.
That means moving comfortably between operating models, process design, architecture, agentic systems, automation, evaluation, governance, and value realization rather than treating them as separate disciplines.
02 / POINT OF VIEW
AI transformation is not a model-selection exercise.
The strongest systems are often the ones that know when not to use an agent, when deterministic automation is enough, when retrieval is the actual requirement, and when humans should remain explicitly in the loop.
I care about architectures that are open, measurable, understandable, and economically defensible.
03 / CURRENT THEMES
Operational agents
Routing, state, tool use, approvals, failure handling, observability, and production evaluation.
MCP-first architecture
Open interfaces, thin orchestration, and fewer proprietary integration dependencies.
Tokenomics
Cost per successful task, retry tax, context inflation, human review, model routing, and payback.
Operating systems for AI
Repeatable intake, architecture, evaluation, governance, adoption, and value realization.