Enterprise production
Python automation, Kubernetes integrations, observability migrations, and distributed-system diagnosis across complex customer environments.
← Recruiter fast track
Platform engineering / AI infrastructure / operations
I build the layer other engineers depend on.
Enterprise production experience, internal developer platform delivery, and a privately operated cloud-native estate converge in one working style: make the platform legible, reusable, observable, and recoverable.
10+
years in production systems
9
k3s nodes operated
25+
GitOps applications
10
CNCF certifications
01 — Why the fit is real
Evidence across employment and independent builds.
Developer platforms
Hands-on adoption work across Kubernetes, CI/CD, identity, APIs, and cloud infrastructure for enterprise internal developer platforms.
Day-two operations
GitOps, networking, identity, secrets, state, metrics, logs, traces, recovery, and upgrades operated as one connected system.
AI infrastructure
Private vLLM inference, GPU capacity, retrieval, memory, agent tools, and the telemetry required to operate them safely.
02 — Direct proof
Four paths. No scavenger hunt.
System
The Hive
Nine-node k3s, Argo CD, Cilium, Longhorn, 102 TB ZFS, and a full observability plane.
Inspect →
Architecture
TechBearing
Approximately fourteen services exposing durable operational context through MCP over SSE.
Inspect →
Operations
Failure Lab
Real incidents and near misses converted into causal chains, recovery gates, and stronger platform controls.
Inspect →
Evidence
Live Systems
A public-safe snapshot of real platform state without creating an access path.
Inspect →
03 — Interview walkthrough
Show the reasoning, not a feature tour.
Ask me to trace a failure across Kubernetes, identity, networking, storage, observability, and an AI workload. I can show the architecture, the evidence path, the recovery decision, and what changed afterward.
01
Start with the user-visible symptom
Establish impact before selecting a familiar layer to blame.
02
Follow evidence across boundaries
Trace network flow, workload state, storage, identity, and telemetry as one request path.
03
Restore, then explain
Separate containment from root correction and make the recovery path executable.
04
Change the operating system
Leave behind a guardrail, alert, runbook, or platform capability that reduces repeat work.
The short version
I do not stop at understanding the system. I ship the capability and stay close enough to learn whether it worked.
Download matching résumé ↓