I'm Chris Rothmeier — an infrastructure & systems engineer in Philadelphia. Thirteen years building, operating, and improving enterprise infrastructure, from Active Directory forests and Exchange to hybrid Azure and a self-hosted, high-availability Kubernetes and GPU datacenter.
My career started in traditional Microsoft infrastructure — Windows Server, Active Directory, VMware — and evolved through hybrid cloud (Azure AD Connect, Conditional Access, Microsoft 365) into Linux administration, Kubernetes, and GPU compute. The constant across all of it: I care about whether systems actually work when it matters most.
I run a self-hosted datacenter not as a hobby but as a validation environment — testing technologies against real constraints (storage, networking, GPU scheduling, observability, recovery) before I rely on them or write them up. The lab notes on this site document that process honestly, including the parts that didn't go to plan.
Windows / VMware estates, identity (AD / Entra ID), M365 administration, endpoint security — day-one ready with 13 years of direct experience.
Infrastructure-as-code, observability, and automation-first operations with a reliability and change-safety mindset.
Multi-client infrastructure management, client-facing communication, and broad stack coverage — a strong fit for Tier 3 MSP roles.
W-2 employment only. Not available for freelance or contract work. Philadelphia area or remote.
Windows Server, Active Directory, VMware vSphere, Exchange — traditional enterprise infrastructure with strict uptime requirements.
Azure, Microsoft 365, hybrid identity, Azure AD Connect, Conditional Access.
Kubernetes, Docker, infrastructure-as-code (Terraform / Ansible), Linux administration.
High-availability K3s, a multi-GPU NVIDIA fleet, local inference workloads, and full observability.
Current stack
Proxmox / KVM · Kubernetes (K3s) · NVIDIA GPUs · ZFS storage · Prometheus / Grafana · Terraform / Ansible.
Lab work is for learning and validation — not a product or a services offering.
A heterogeneous multi-GPU NVIDIA fleet spanning datacenter and current-generation workstation cards, used for local inference and GPU-scheduling validation, with per-GPU scheduling and monitoring.
10GbE backbone with 25GbE point-to-point between Proxmox hosts. Multi-VLAN segmentation, QLogic adapters, DAC cabling, and proper isolation for lab vs. production traffic.
ZFS-backed pools with snapshots, replication, and NVMe caching. Proxmox Backup Server for automated VM / container backups with verified restores.
Prometheus + Grafana with custom dashboards, alerting tied to real failure modes, and a DCGM exporter for GPU metrics. If it isn't monitored, it doesn't exist.
Open to full-time W-2 infrastructure or systems engineering roles — Philadelphia, PA or remote.
