I’m a senior IT leader and hands-on systems integrator with 23 years in the field, all of them at one managed service provider. I started at entry level and grew into my current role as Senior Systems Integrator and Director of Information Technology at Mnemonic Computer Solutions. Mnemonic is a fully remote team today, with its business address in Oak Park.
I’m the person people come to when a problem needs critical thinking, from the help desk queue all the way up to the CEO. I work across the full stack: network architecture and firewalls, identity and Microsoft 365, cloud, backup and disaster recovery, file platforms and endpoints. I’m also the escalation point for the hardest tickets.
Every technical decision at Mnemonic is made in consultation with me, and the CEO brings me into the decisions that shape the company: technology and business direction.
What makes my role a little unusual is that I build things too. I design and ship production MCP servers, AI agent skills and n8n pipelines, and every year I lead a large live-event IT build from a blank page to a working show floor.
Working style
- Root cause first. I want to know why something happened, not just make the alert go away.
- Question the assumptions. Understanding why a system is built the way it is usually reveals the real fix.
- Document everything. Runbooks, SOPs and a working knowledge base are how a small team scales.
- Automate what’s understood. Once a problem is understood end to end, it becomes a candidate for a script, a pipeline or an AI agent.
Outside work
I love building and refining things. I’ll take something apart to understand how it works, question the design, and look for ways to improve it. That curiosity follows me outside work, especially into sim racing.
My sim rig is a project I keep evolving. I’ve designed custom parts and accessory mounts, built a 3D-printed button box with an integrated Stream Deck, and made a button enclosure for my Porsche replica steering wheel. I also designed how the wheel’s controls connect to my other button box and make my own cables for different applications, including speakON connections for tactile feedback.
The rig brings together mechanical design, electronics, wiring and software: a custom dash, a modified chassis, and a custom seat plate that mounts the seat, isolators and tactile hardware. I enjoy figuring out how the pieces fit together, testing an idea, and refining it until it works the way I want.
Explore my sim rig and hardware
About this site
This site doubles as my AI lab. AI agents (Claude, ChatGPT and Codex, OpenCode, and whichever harness I’m testing that week) build and maintain it, and every one of them follows the same written rulebook. Here’s how it fits together.
- Staging. A private copy runs in Docker on openclaw01, my own Linux box. It’s only reachable over my Tailscale network and never faces the internet. Every change happens there first.
- Production. The site you’re reading is hosted on Hostinger, with DNS and the CDN handled by Cloudflare.
- Deploys. When a change is ready, one script pushes it from staging to production over SSH. It snapshots both sides, syncs files and the database, clears the cache and checks the site is up. If anything looks wrong, one command rolls production back.
- Safety net. A snapshot is taken before every change, on staging and in production. Images are stored once and shared between snapshots, so backups stay small.
- Paper trail. Every change is committed to a private GitHub repo under the name of the agent that made it, and every working session ends with a public Lab Note.
You can read what worked and what didn’t in the lab notes, and see how the agents compare on the Harness Scoreboard.