Sysadmin · Developer · Infrastructure Enthusiast
Building robust systems & elegant code
ExploreI'm Patrick, known online as chimme — Head of IT running infrastructure across 20+ medical facilities, turning complex healthcare environments into well-oiled machines.
I architect and manage everything from Hyper-V clusters and SQL Server estates to custom-built APIs, AI-powered phone systems, and automated lab pipelines. If it touches a network, a database, or a deployment pipeline — it's probably mine.
I build the tools I wish existed: patient document archives that actually work, lab result systems with AI-driven assessments, and infrastructure that scales without drama. The best system is the one nobody notices — until you see what it handles.
The tools and technologies I work with daily
A selection of things I've built and contributed to
Multi-agent AI system with 5+ autonomous agents handling ops, monitoring, coding, and recovery across distributed infrastructure — 24/7, fully self-healing.
Patient document archive serving 2.5M+ documents across 20+ medical facilities. Deep database integration, automated categorization, and seamless EHR connectivity.
Automated lab result pipeline ingesting LDT 3.0 data from external labs, with AI-powered clinical assessments and automatic patient record matching.
Complete IT infrastructure for 20+ cardiology and pulmonology locations — Hyper-V clusters, failover clustering, VPN mesh, and centralized management.
Internal tools for tracking software licenses across all locations and managing vendor invoices — replacing spreadsheet chaos with clean dashboards.
Full observability across the entire fleet — uptime monitoring, automated alerts, health dashboards, and self-healing recovery agents.
...and many more
Running production healthcare IT across 20+ locations — where downtime isn't an option.
Multi-node Hyper-V clusters with live migration, failover clustering, and SQL Server Always On — keeping critical medical systems running 24/7.
WatchGuard firewalls, WireGuard VPN mesh across all locations, Tailscale for remote management, and centralized DNS — one network, zero excuses.
Docker containers, CI/CD pipelines, and AI agents that deploy, monitor, and self-heal — most issues get fixed before anyone notices.
Automated nightly backups with offsite replication, tested restore procedures, and AI-driven recovery agents on standby for instant response.
I don't use AI. I build AI systems that run production infrastructure — autonomously.
While most people are prompting chatbots, I'm operating a fleet of AI agents that ship code, heal servers, analyze medical data, and coordinate with each other — 24/7, no human in the loop. This is what AI looks like when you stop playing with it and start building with it.
5+ specialized AI agents with persistent memory, cross-agent communication, and independent decision-making. They delegate tasks to each other, spawn sub-agents for complex work, and operate across Discord, WhatsApp, and internal APIs — around the clock.
Fully automated lab result pipeline: ingest raw LDT 3.0 data, match patients via Oracle DB, then run AI-powered clinical assessments cross-referencing ICD-10 history — delivering actionable insights to physicians before they even open the chart.
Recovery agents on dedicated hardware that detect failures, diagnose root causes, SSH into affected systems, and execute fixes — all without human intervention. Mean time to recovery: minutes, not hours. Most incidents resolve before anyone notices.
Coding agents that take a task description, explore the codebase, implement features, write tests, open PRs, and deploy to production. Full cycle from idea to shipped code — sometimes faster than I could explain the problem to a human developer.
Latest from the frontier — curated and updated automatically.
Interested in working together or just want to say hi?
