Profile / AI Infrastructure

Infrastructure for intelligence at scale.

I lead the programs that turn accelerators, Linux, high-performance networking, and fleet engineering into production AI platforms.

Read my profile Field notes on Substack
Silicon to fleet 20+ years in critical infrastructure 10+ patents
01
The infrastructure stack

AI performance is a system property.

Reliable AI platforms emerge when every layer is designed, qualified, and operated as one system.

01 / Silicon

Accelerators

GPU, custom silicon, and server platforms

02 / Systems

Fabric

HPC networking, storage, and virtualization

03 / Software

Linux

Images, kernels, drivers, and frameworks

04 / Operations

Fleet

Validation, reliability, and production learning

02
How I lead

Make the whole system ready.

I translate workload and customer requirements into technical plans, expose cross-stack dependencies early, and create the mechanisms that carry complex infrastructure from definition to fleet adoption.

Experience and operating philosophy
  1. Define

    Product outcome, architecture, requirements, and decision points

  2. Integrate

    Hardware, firmware, Linux, networking, storage, and validation

  3. Qualify

    Performance, reliability, risk, and production readiness

  4. Learn

    Fleet telemetry, incident signals, and durable engineering improvements

03
Field notes

Writing from inside the infrastructure stack.

AI systems, Linux, cloud platforms, engineering quality, and the work of leading at scale.

Read on Substack