How I lead
I work at the intersection of product architecture and organizational design. My background spans robotics, automotive systems, connected products, autonomy compute, payment hardware, and commercial-fleet technology.
Across those settings, I have become interested in a consistent question: how do I help a group of capable people produce a result that is stronger than a collection of individually completed tasks?
Give teams an outcome they can reason about
I want people to understand the problem, the quality bar, and the constraints. A detailed implementation prescription can create activity while leaving the underlying purpose unclear. An observable outcome gives a team a basis for making its own decisions and bringing useful alternatives back into the conversation.
Direction needs to survive implementation. I try to make it visible in priorities, acceptance criteria, architecture, and review rather than relying on repeated clarification from me.
Own the complete behavior
My current work at Samsara spans Safety and Telematics, including firmware, edge AI, quality, automation, and supporting services and data. That breadth is useful when it reduces gaps in product ownership.
I want teams to understand the contracts between their systems and have a practical way to resolve cross-functional tradeoffs. Hardware constraints, model behavior, testability, field operation, and backend cost all belong in the discussion when they affect the same outcome.
Build feedback into the work
Fast implementation helps when the system can establish what changed and whether it is acceptable. I invest in tools, observability, reproducible environments, and tests that make important information available close to the decision.
That is also how I think about agentic development. The model’s capability is one input. The surrounding workflow determines how much of that capability becomes useful progress. I want to measure accepted results, reliability, and rework rather than assume that more generated material is an improvement.
Make autonomy practical
Clear ownership and escalation paths let people act without bringing every reversible choice through a central queue. Shared definitions and durable decisions reduce the context they need to reconstruct.
My responsibility is to supply direction, resolve consequential conflicts, and help develop judgment. I want managers and engineers to own their work, with enough support to grow into a broader scope.
Keep the organization able to do good work
Changes in structure and priorities affect trust, relationships, and people’s sense of where they can succeed. I treat continuity and clarity through those changes as part of delivery leadership.
AI agents can increase execution capacity. They do not remove the responsibility to develop people, communicate honestly, and build a team that can sustain its progress.