Infrastructure and DevOps professional with about a decade of experience building, automating, and operating technology environments. Operates a production Linux platform with 100+ containerized services and more than 100 users. Brings hands-on experience in Docker, networking, monitoring, incident recovery, APIs, and automation in Python, Bash, PowerShell, and JavaScript. B.S. in Computer Science.
Infrastructure and DevOps professional with about a decade of experience building, automating, and operating technology environments. Operates a production Linux platform with 100+ containerized services and more than 100 users. Brings hands-on experience in Docker, networking, monitoring, incident recovery, APIs, and automation in Python, Bash, PowerShell, and JavaScript. B.S. in Computer Science.
Grouped by where it belongs in the lifecycle. Everything here is in production on infrastructure I run, not a checklist padded to look full.
About a decade of keeping people and systems running, from the help desk to the server room.
Everything below runs in production. It has real users and uptime I'm accountable for, and I've had to dig it out of real failures. I built it, deployed it, and still run it.
A multi-service, containerized platform behind one hardened edge. I own every layer: host provisioning, container orchestration, reverse-proxy routing, automated TLS renewal, intrusion prevention, dynamic DNS, backups, and incident recovery.
Problem: hold dozens of real-time transcodes at once on consumer hardware, behind a residential uplink, without causing buffering or saturating my upstream.
What I own: a multi-backend streaming tier with hardware-accelerated transcoding, distributed sessions, bandwidth shaping, usage telemetry, and capacity planning that balances workloads across available compute while sustaining reliable service.
Outcome: used AI-assisted analysis and agent workflows to reverse engineer software behavior, then turn those findings into working tools and integrations.
What I built: an AI development platform across Linux and Windows, with selectable text and vision models, a C# service-control app, MCP integrations, and guarded agents for research, implementation, and review.
Problem: give tenants isolated, remotely managed compute and file storage on shared hardware.
What I own: VM provisioning and isolation, the storage layout, centralized remote management, and new tenant setup.
Problem: running the platform solo leaves me with recurring chores that drift if I ignore them: a rotating IP, DNS chasing it, databases too slow on disk, and media metadata that's wrong out of the box.
What I own: the automation that handles those chores. Databases run from a RAM disk, with tmpfs bind-mounted over the container and a nightly backup that pauses the service only for a sub-second RAM-to-RAM copy. The remaining automation covers IP tracking, DNS reconciliation, fail2ban operations, batch metadata edits, and registrar automation.
I build, operate, and recover infrastructure with hands-on ownership of the systems behind the service.