Tech Projects

Networking, automation, AI, and hands-on engineering. 17+ years of building things that work

Homelab & NetDevOps

Running a home lab since 2002 — today it’s a full enterprise analog: OPNsense firewall, Cisco switching, NetBox as the network source of truth, Ansible orchestrated through Semaphore, and a Grafana + Telegraf + InfluxDB monitoring stack, all running on Docker and self-hosted GitLab with GitHub mirroring. The newest layer is local AI: LM Studio serving models on consumer GPU hardware, with agentic coding tools that write and test Ansible playbooks, push them through GitLab CI, and deploy to real network gear. It’s a production-style NetDevOps pipeline that fits in a home office.

Read: One Shared Brain for All My AI Assistants →

Automation, Scripting & Tools

Python and Ansible are the daily drivers. Highlights include Ansible roles for Cisco config pushes with vault-secured credentials and Jinja2 templating, a Flask-based IPAM tool that uses Python as the brain and Ansible as the hands, web scrapers that turn junkyard inventory into structured data, and a full stock-screening and backtesting platform with broker integration for paper trading. If a tool doesn’t exist, I build it.

Networking, CAD & Other Tech

Random work that may be interesting enough to post here — I’m always working on something. Network gear, 3D-printed brackets and shop tools designed in CAD, and whatever else the current project needs.

Off-Grid Solar

Designed and installed a 48V LiFePO4 off-grid solar system to code — 30psf and 160mph wind rating, fused, surge-protected, and grounded, with battery capacity comparable to a Tesla Powerwall at a fraction of the cost. Engineering isn’t just racks and code.

Write-ups

The build logs behind the projects above, newest first.

  • Isolated Enough to Be Safe, Integrated Enough to Be Useful
    The controls between my AI agents and the systems they touch: VLANs, unprivileged containers, scoped AppRole tokens, contributor-only Git access, mandatory secret scans, and fixed automation templates.
  • Keeping AI Coding Agents Moving on Multi-Step Projects
    The useful part of an AI coding agent is not the first hour. It is whether the work still has direction after the context fills up. A few plain files and clear verification gates make the difference.
  • The Restore Is the Backup
    A backup job that finishes is only the first link. Recoverability depends on consistent capture, validation, retention, an independent copy, and a restore test.
  • SSH Key Authentication Without Locking Yourself Out
    SSH keys are straightforward once authentication and authorization are treated as separate problems. Add access carefully and keep a way back in.
  • Proxmox: What the Host Actually Runs Now
    The migration was the easy part. Year two is templates, Vault, self-hosted CI, and a second network tier. That is the accretion that turns a hypervisor into a platform.
  • Enterprise Containerlab Setups
    How route reflectors, EVPN/VXLAN, dual transport, and CIN models work, why enterprises use them, and when simpler designs may be better.