About
I build the automation your team keeps meaning to get to.
Two decades in infrastructure and platform engineering. The last few years spent working out where AI actually earns its place in an engineering workflow — and where it doesn't.
What Blue Ridge Systems is
Blue Ridge Systems is one person. No account managers, no bench of juniors handed your work after the first call. You get two decades of infrastructure and AI systems experience on your problem, start to finish.
That's deliberate. The value isn't headcount — it's judgment. Knowing when AI is the right tool and when a queue or a deleted process would serve you better. Knowing how to scope work so it actually ships. Knowing what enterprise-scale systems look like from the inside, and how to apply that to teams that don't have enterprise-sized budgets.
I founded BRS in late 2024 on a specific thesis: the best applied-AI work starts with someone who understands the infrastructure underneath, not just the models on top. Most of what breaks in an AI deployment isn't the AI. It's everything around it — the data plumbing, the failure modes, the operational load nobody scoped.
Work worth judging me on
Apstra CloudLabs
2017 – 2025Platform · CI/CD
Took a handful of internal Ansible scripts and rebuilt them into a global, self-service lab platform — dozens of bare metal instances across multiple global regions, wired into engineering's own CI/CD so sales always demoed the current build. Survived the Juniper acquisition and became the platform that onboarded a Fortune 500 sales organization. Operated by four to five people. Still running.
Agentic development, proven on my own builds
2024 – presentMethodology
Two production systems built with AI executing under my architectural direction: a full Python/FastAPI SaaS written without hand-coding, and Root Cause — a simulation in Rust and Bevy, a stack I had never used, shipped with 80+ integration tests. My own projects, not client work, and I am explicit about that. The methodology is what transfers.
apstra-api-python
2019 – presentOpen source
Open-source Python client for the Apstra REST API, written to move network engineers from CLI habits into API-driven automation. In continuous use since 2019.
Embedded AI discovery — major league sports organization
2025Embedded
Six months embedded with a lean DevOps team, working out where AI could accelerate their infrastructure automation and observability. The organization never funded the initiative, so it stayed at discovery — worth saying plainly, because the lesson shaped how I scope this work now: find out whether the mandate has a budget behind it before anyone commits months.
How I got here
I grew up in Amarillo, Texas. Hockey was the plan until it wasn't. At nineteen I walked into a Cisco networking class at Amarillo College and walked out with my head on fire. I'd barely touched a computer before that.
The path wasn't straight — cybersecurity, an MBA, then back to networking because I wanted to work on systems, not write policy. I spent a decade grinding toward the CCIE while the industry shifted underneath me: watched SDN fail loudly, watched network automation find its legs as DevOps ideas finally reached infrastructure. PNC gave me the scope to chase it, and I stood up their network automation function.
At Apstra I got properly humbled. I arrived as a network engineer who barely knew what an API was and got kicked into shape fast. Building CloudLabs taught me the thing I still lead with: infrastructure doesn't have to be hard to operate — it has to be designed right.
My route into AI looked exactly like my route into networking. I woke up early on a Saturday because I couldn't wait to get back to it. Right after ChatGPT shipped I showed the Apstra team what I'd been building, over lunch, and watched the lightbulbs go on around the table. That's the gap I work in: most teams still haven't seen what this can do on their own problems.
I left Juniper in 2025 to do this full time. The work I want is helping engineering teams find the gaps they didn't know to look for, building the systems that close them, and making AI something your team controls rather than something happening to it.
Credentials
Want to talk through a problem?
Bring the messy version. I'd rather hear the actual constraint than a cleaned-up brief.
Book a Call