AI infrastructure, from silicon to fleet

I build the infrastructure that AI depends on.

I am an AI infrastructure and technical program leader with more than two decades of experience bringing complex hardware and software platforms from early definition through production deployment. My work sits at the intersection of accelerated compute, hyperscale cloud systems, Linux, high-performance networking, and the AI workloads they support.

As a Principal Program Manager in Microsoft Azure Compute, I lead cross-functional programs for GPU and AI accelerator platforms. I translate workload and customer requirements into technical plans, align dependencies across hardware and software teams, manage readiness risk, and drive integrated systems through qualification, production sign-off, and fleet adoption.

What I Lead

• AI and accelerated compute: End-to-end readiness for GPU, custom accelerator, HPC, bare-metal, confidential computing, and ARM64 platforms.
• Integrated systems: Alignment across silicon, firmware, virtualization, Linux images, kernels, drivers, networking, storage, and platform validation.
• High-performance networking: Linux enablement for MANA, Accelerated Networking, DPDK, InfiniBand, SR-IOV, and the demanding scale characteristics of AI clusters.
• Quality and reliability: Validation strategy, production quality gates, telemetry, incident learning, and automation that improve platform readiness and fleet health.
• Partner and customer leadership: Technical alignment with engineering organizations, silicon and Linux ecosystem partners, and customers deploying mission-critical workloads.

Building Platforms, Not Just Features

AI infrastructure succeeds only when the full system works together. I focus on the seams between components, where issues in firmware, drivers, kernels, networking, storage, or orchestration can become workload-level failures. That systems view helps teams make better tradeoffs across performance, reliability, operability, and time to market.

I have built platform-quality programs and validation services from the ground up, led Linux and driver readiness for new compute generations, and developed AI-assisted approaches to test selection, failure triage, and release decisions. The goal is consistent: turn complex integration work into a repeatable engineering system that can scale.

From Infrastructure to AI Workloads

I also work hands-on with modern AI systems, including large language model inference, serving frameworks, performance benchmarking, quantization, and agentic workflows. Understanding prefill, decode, concurrency, memory behavior, and failure modes connects infrastructure decisions to the outcomes that matter for AI developers and customers.

Career Perspective

Before my current role, I led Linux kernel productization and infrastructure programs across Azure and partner platforms. Earlier at T-Mobile, I directed national capacity and network evolution programs and applied predictive machine learning to large-scale infrastructure planning. I am also an inventor on more than ten patents spanning machine learning, network quality, and infrastructure intelligence.

How I Lead

I connect the technical and business worlds: defining the product outcome, making system dependencies visible, creating clear decision points, and keeping teams focused on measurable customer impact. I value technical depth, evidence, direct communication, and durable mechanisms over one-time heroics.

Outside of work, I am a mountaineer and ski instructor. Both reinforce the same leadership habits I rely on in technology: preparation, respect for constraints, and calm decision-making when conditions change.