Ask Me About
Conversations I welcome. If your question lives in one of these, I'd love to hear from you.
Multi-agent system design
Orchestration patterns, when agents should disagree, evaluation harnesses.
- How do you decide where to put guardrails vs. let agents reason?
- How do you evaluate a multi-agent system?
AI in regulated workflows
Banking, healthcare, prior authorization, claims — where AI can land safely.
- What breaks first when you put an LLM into a regulated workflow?
- How do you structure human-in-the-loop without killing throughput?
Architecture at altitude
Moving cleanly between board-level intent and code-level decisions.
- How do you translate strategy into architecture without lossy summaries?
- What architecture decisions are board-relevant?