My technical foundation is real: six years as a USAF-trained weather forecaster (distinguished graduate) working observation networks and satellite systems and writing the SOPs for deploying them, then backup administrator at Constant Contact — enterprise storage and server installs in co-located datacenters, tape backup administration, an LTO3→LTO4 migration. After that I left tech for over a decade and worked deliberately varied, human-scale jobs: farm management, caregiving, facilities maintenance, a massage practice. In 2025–2026 I came back through the AI door, and I now design, run, and maintain a production-grade environment end to end, daily, using AI agents as my workforce and myself as the architect, reviewer, and incident commander.
I'm not a senior engineer and I won't pretend to be one on a whiteboard. What I am is an operator who ships: I own a fleet of real services with real users, real backups, real monitoring, and a real incident history — and I run it at a speed one person couldn't reach without knowing exactly how to direct AI and exactly when not to trust it.
what I run (the receipts)
One Linux server + two satellite machines, operated as production. Everything below is live, versioned, and documented.
Service fleet: 8 public domains behind nginx + Cloudflare Tunnel, 10+ systemd services, 25+ scheduled jobs (data pipelines, scrapers, digest emails, dashboards), all version-controlled including host config and crontab snapshots.
Backups that restore: nightly rclone to Cloudflare R2 across three machines (Linux, macOS, Windows), with failure alerting, preflight checks, 60-day retention lifecycle — and tested restores, including a real recovery of a wiped crontab.
Security posture: tiered exposure (Tailscale-only vs LAN vs public), ufw, daily CVE and dependency scans, nightly webroot-exposure scan born from a real incident I found, fixed, and turned into a guardrail.
Monitoring: uptime checks with SSL-expiry alerting on every public site, log review pipelines, reboot-required checks across machines.
AI orchestration as an ops discipline: I run a documented delegation protocol — which model/agent gets which task class, what requires human review, what counts as a machine-checkable done-test. Published tooling: claude-dmg-skill (docs/memory/git sync skill for Claude Code, posted to Hacker News).
Builds: Python pipelines (market-data collectors with replay-deterministic engines), Flask services, scrape-diff-notify monitors, an EEG biofeedback system, email automation. Spec'd by me, drafted by AI, reviewed and verified by me.
what that's worth to you
Small IT team or MSP: one person who can own the unglamorous whole — backups, monitoring, patching, DNS, migrations, documentation — and who automates himself out of repetitive work by default.
Ops/NOC: calm, methodical incident behavior: evidence before action, one variable at a time, rollback path stated before any state-changing fix. I keep written playbooks because I've paid for every line in them.
Hardware-adjacent roles: datacenter and break/fix background — racking, storage, the physical layer most AI-era candidates have never touched.
Healthcare IT: the care thread in my history is long and deliberate — certified Home Care Aide, direct-care work with disabled adults, CPR/First Aid, and EMT-B training starting Aug 2026. If you need someone who speaks both systems and patient-care, that intersection is where I'm headed on purpose.
The honest part: my fluency is AI-augmented. Unassisted, I'm a competent journeyman, not a wizard — at a raw terminal with no copilot I'll be slower than your senior engineer. But the augmented version of me is who shows up to work every day, the judgment about when the machine is wrong is entirely mine, and that judgment is the actual skill now. If your shop bans AI tooling, I'm the wrong hire. If your shop is trying to figure out how to use it safely, I'm a cheap way to find out.
history
2025–now
Independent systems operator — the environment above, run as production daily
2011–2024
Deliberately varied decade: farm management (CSA farm, TN; 1,200-bird egg operation, WA), certified Home Care Aide and direct-care work, facilities/pool maintenance (Ojo Caliente Spa, NM), licensed massage therapist, ops for a toy startup
2010–2011
Constant Contact, Waltham MA — Backup Administrator: enterprise server + storage installs in co-located datacenters, tape backup administration, LTO3→LTO4 migration
2001–2007
US Air Force / Air National Guard — Staff Sergeant, weather forecaster (distinguished graduate): satellite + field observation systems, networked station operations, authored deployment SOPs. Honorably discharged.
upcoming
EMT-B, San Juan College — begins Aug 2026
stack
Linux (Debian/Ubuntu/Mint)bashPythonnginxsystemdcrongitCloudflare (DNS/Tunnel/R2)TailscalercloneFlaskSQLiteSMTP/email automationClaude Code / Codex CLI orchestrationWindows + macOS fleet care