This is where I document what I’m learning and building in AI infrastructure — in the open, with measurements.
- Concepts — evergreen explanations. One idea, clearly explained, with the enterprise angle.
- Articles — deeper technical investigations, backed by measurements.
- Lab — what I’m building, what breaks, and what the numbers say.
About
I’ve spent my career building and operating infrastructure where reliability isn’t optional.
IBM Power, x86, HPE Superdome, SAP HANA on bare metal, enterprise storage, SAN fabrics, and high-availability systems.
I’ve worked across hardware, firmware, operating systems, networking, storage, automation, and performance — including the problems that only show up when a system is supposed to be working.
I lead, design, and build infrastructure using custom system deployment tools, BMC/RMC, Dell iDRAC, HPE iLO, Redfish/API automation, system firmware update automation, performance monitoring, observability, and purpose-built solutions — applying enterprise reliability and performance engineering to AI infrastructure.
Now I’m applying that experience to GPU platforms, Kubernetes, inference serving, and observability.
I’m documenting the build as I go — what works, what doesn’t, and what I learn along the way.
How I Think About Infrastructure
Good infrastructure isn’t just about getting a system running. It needs to be understandable, observable, repeatable, and recoverable. Performance matters. Reliability matters. Cost matters. And eventually, someone has to operate the thing. That’s the perspective I’m bringing to AI infrastructure.
Engineering Focus
I’m continuing to build deeper expertise in GPU platforms, Kubernetes, inference, observability, and AI infrastructure. My goal is to connect hands-on engineering with architecture, operational discipline, and business outcomes.
Contact
I document what I’m learning and building in AI infrastructure — follow along if you’re working on the same problems.
- LinkedIn: profile — connect with me professionally.
- GitHub: ai-infrastructure-lab — see what I’m building and experimenting with.