Senior Platform Engineer at Pierce
From the physical rack to AI operations.
I follow the problem through every layer, then leave behind a system other people can run.
Every useful infrastructure question starts the same way: what is actually connected to what?
I traced the estate and modelled it in NetBox. VLAN by VLAN. Cable by cable. Port by port.
Tracing beats transcribing: a model is only worth having if it is complete. I made it something any engineer can query, with configuration backups behind the switch fleet.
The estate can be reasoned about
before it is touched.
Paths traced, changes reviewed, improvements proposed from the model rather than from memory.
I own the caching rules at the edge — what gets cached, for whom, and for how long, across older parts of the stack carried through a platform migration. What makes them worth real attention is that a status code tells you nothing about what was actually served: the same URL can come back full or light depending on the requesting agent and whether the cache is warm. So I read response bodies, per agent, warm and cold.
Conceptual response comparison. No production page or response body is reproduced.
Read what the server actually returned. Compare the bodies, not just the status codes or keyword dashboards.
Cache behaviour is infrastructure, and it decides what the outside world actually sees. Caching is where an infrastructure setting turns into a commercial outcome, so it is worth one person owning it deliberately — especially while a stack is mid-migration and two generations of it are live at once.
The same pattern applied to packages and patching: Red Hat Satellite and Pulp as the source of truth for what exists on which machine, and a defined procedure for getting security updates onto the fleet.
A patch round is a repeatable operation with an auditable result: content versioned deliberately, promoted in order, and answerable per machine. The platform upgrade and the readiness for the next operating systems came out of the same discipline.
The test is what remains after the change: known state, repeatable operations and fewer things that depend on memory.
Large monitoring dataset recovered and kept healthy by database tuning.
PierceCertificate lifecycle automated end to end: issue, store, renew, deploy.
PierceSole infrastructure on-call through Black Week, seven days straight.
Pierce · 2024 & 2025Migrated from Akamai to Cloudflare, including edge routing and Terraform reconciliation.
Pierce · 2026Depth at one layer helps. Following a problem across the boundaries is where I do my best work.
Internal LLM service deployed company-wide.
INDEPENDENT INFRASTRUCTUREMulti-agent operations, persistent memory and scheduled health rounds.
AWS organization built from the root account up. Entra ID SCIM, dual-active IPSec design, Azure and Cloudflare Workers.
Docker and Harbor. Harvester 1.7 / RKE2 proof of concept on nested vSphere. API-gateway evaluation and ADR: Kong against KrakenD.
INDEPENDENT INFRASTRUCTURETalos Kubernetes, Cilium and Kube-OVN.
Zabbix 7.0 LTS with automated proxy deployment. ELK major upgrade, Grafana, Kafka consumer lag via Burrow and node_exporter.
INDEPENDENT INFRASTRUCTUREVictoriaMetrics.
Akamai to Cloudflare migration, Workers routing and Terraform state reconciled with production. SPF, DKIM and DMARC across all domains, with SPF flattening.
NetBox modelling, switch configuration backups, firewall changes by API, VLANs and IPSec/VPN.
INDEPENDENT INFRASTRUCTUREVyOS routing.
NetApp NFS/CIFS. Blade-controller latency traced to NFS nconnect tuning.
INDEPENDENT INFRASTRUCTURELINSTOR / DRBD and ZFS.
VMware / ESXi on Cisco UCS, Linux, Red Hat Satellite and Pulp.
INDEPENDENT INFRASTRUCTUREProxmox.
The secrets and certificate platform follows that rule: deployments pull secrets at deploy time; certificates renew and deploy themselves. Rotation becomes a repeatable process, rather than a favour asked of the person who built it.
Different constraints. The same engineering standards.
Talos Kubernetes on Proxmox. Cilium and Kube-OVN networking, LINSTOR/DRBD and ZFS storage, VyOS routing.
The constraint is one person and no rota. The operations workflow is designed to turn collected state into a decision about what needs human attention.
A Telegram-first inventory and listing assistant. Photograph an item, get recognition and a ready-to-post listing, publish to marketplaces.
Real users, Stripe billing and eBay integration. Python, Telegram Bot API, PostHog and Grafana.
A Ukrainian biology and ecology portal. Migrated from Drupal to Django and Wagtail through a rehearsed cutover, with no acceptable downtime window.
Visit BatrachosA kit and upkeep companion for cyclists and runners. Photo-based gear cataloguing, weather-aware clothing and component wear from ride data.
A working alpha. An ongoing product, not a finished release.