Systems Operator. Architect.
I work inside one of the hardest engineering environments to fake: enterprise payments. At The Home Depot, I contribute to a tender ecosystem that supports 2400+ stores, high-throughput transaction flows, and platinum-tier uptime expectations.
My focus is not generic full-stack work. I instrument critical paths, reduce operational ambiguity, and help move risky systems without customer impact. That is the core skill behind my career shift toward staff and principal-scope systems architecture.
The next layer is hardware-cloud fusion: ESP32 nodes, Raspberry Pi 5 gateways, machine vision, and GCP analytics tied into public observability boards. I care about the whole chain, from noisy sensors and power budgets to trace-linked cloud events and operator tooling.
What I Build
- Instrument Critical Paths
I build the telemetry that turns vague production pain into measurable facts. Traces, dashboards, alert rules, and canonical logs are architecture, not afterthoughts.
- Move Risky Systems Safely
Zero-downtime migrations, rollback gates, traffic shadowing, and tight change discipline are the difference between architecture theater and operating systems that survive contact with reality.
- Build Edge-to-Cloud AI Systems
I am less interested in chatbot wrappers than in AI control planes: local inference, retrieval-backed runbooks, public telemetry, and hardware that reports into the cloud as a coherent system.
Operating Principles
Measure before opinion
If a system is important, it deserves traces, metrics, and an incident timeline. Guesswork is not an operating model.
Design for failure, not the demo
Rollback paths, cache expiry, replay protection, and error budgets matter more than pretty architecture diagrams.
Use AI where operators need leverage
AI is valuable when it summarizes incidents, retrieves runbooks, or helps reason about edge fleets. It is not valuable as decorative product frosting.