Now

Last updated: September 2026

Building

I am currently focused on edge AI systems where inference quality, latency, memory budget, and runtime behavior all need to be balanced on constrained hardware.

Learning

• Hardware-aware compilation for AI inference.
• World models and VLA deployment constraints on edge systems.
• Robotics middleware and real-time communication tradeoffs.

Writing

• Notes on robotics architecture and system boundaries.
• Practical observations from inference/runtime implementation.
• Essays on leadership and engineering judgment.

Reading

• Technical papers on embodied AI and robotics perception.
• Writing on compilers and accelerator software stacks.
• Books that surface durable principles for building teams and systems.

Thinking about

How should we design intelligent machines that remain robust when sensing is noisy, compute is limited, and the environment is adversarial?