I help enterprise organizations navigate complex technology problems — from cloud migration strategy and AI governance to full-stack agentic systems — and I don't stop at the recommendation. I build the thing, deploy it, and operate it in production.
Thirty years of infrastructure experience spans on-premises datacenters, virtualization, networking, storage, enterprise software, and cloud platforms — long before AI entered the picture. That foundation is what makes the AI work grounded: every architecture decision is evaluated against operational reality, not just capability demos.
The 30 years isn't a marketing number — it covers the full arc of enterprise infrastructure: on-premises datacenters, storage area networks, virtualization with VMware and Hyper-V, Windows and Linux server environments, enterprise networking, and the first wave of cloud migration starting in the mid-2010s. AI and cloud-native architectures are the most recent chapter, not the whole story.
Most of that career has been in enterprise technology consulting — helping large organizations answer the hardest infrastructure question: what do we actually do with what we have? Cloud migration strategy, application rationalization, vendor selection, build-vs-buy analysis, and the organizational change that makes any of it stick. That consulting background is the lens through which I approach AI: not "what's the most technically interesting system" but "what solves the actual problem, at what cost, with what governance, and who runs it after I leave."
Since 2024 I've gone deep on applied AI — as a builder, not an observer. The Park Whisperer is my personal production AI platform: real-time theme park intelligence running 24/7 across Azure, GCP, and AWS, processing satellite weather data, training ML models, running a multi-agent knowledge RAG, and publishing daily AI-generated content to three social platforms — all operated solo, under $100/month. It's the environment where I stress-test every architectural pattern before recommending it to clients.
Before a model is called, I've already decided: what context is strictly necessary, what can be cached or pre-computed, which model tier matches the actual reasoning demand, and what value the output delivers per dollar spent. That trade-off is explicit in the architecture — not discovered after the bill arrives.