Fritz Anuth
DevOps & Platform Engineer
I build and maintain the platforms that let engineering teams ship faster — from bare-metal to production, from CI to observability.
About
I moved from software development into platform engineering deliberately — fixing the delivery bottleneck turned out to be more impactful than optimising on one side of it. Now I focus on building and maintaining the platforms that let engineering teams ship confidently.
Seven years of hands-on software development inform everything I build on the platform side. I know what developers need and expect from their infrastructure — and I have the coding background to build the tooling, scripts, and automation that actually fits how they work.
My focus is on maintainability and good developer experience, which to me means keeping the platform surface small enough that anyone on the team can orient themselves quickly without asking for help. GitOps fits that philosophy well: version control for everything, infrastructure included.
I work well across organisational boundaries — I've presented migration plans and infrastructure trade-offs to non-technical stakeholders, run postmortems with development teams, and pushed back on decisions when it mattered. I prefer on-site or hybrid working and find being in the room with the team makes a real difference.
My production experience is on-premises and self-managed. Adapting to new environments is part of how I work — I've migrated platforms mid-flight, picked up Java production work alongside an existing .NET stack, and rebuilt pipelines from scratch. I attend KubeCon to stay close to where the ecosystem is heading.
Projects
Homelab Gigawave
3-node bare-metal Kubernetes HA cluster
A highly available Kubernetes cluster built from the ground up on three bare-metal nodes, optimised for power efficiency. Hosts this portfolio, a Forgejo instance, and various self-hosted services — a production-grade lab for testing tooling before deploying it professionally.
Talos Linux as the node OS: immutable, API-driven, no SSH surface. Flux for pull-based GitOps reconciliation, keeping all cluster state version-controlled and auditable. Garage for self-hosted S3-compatible object storage — a deliberate choice to avoid MinIO's licensing drift and stay close to the AWS S3 API.
Running cluster upgrades means draining nodes, testing Flux reconciliation through the disruption, and verifying that the etcd quorum survives the process. This is the environment where new tooling gets evaluated before it reaches production.
Kubernetes
Talos
Flux
Garage
Grafana
Traffic Management & Waiting-Room Service
SP Consult AG
Custom reverse-proxy layer for retail traffic spikes
Designed and built a traffic-management service to protect consumer-facing retail endpoints during sudden demand spikes. The service acts as a configurable admission layer: it queues excess requests and admits them at a controlled rate, presenting a waiting-room experience to end users rather than returning errors or degrading the backend.
Built to handle the kind of traffic pattern common in promotional campaigns — predictable buildup followed by sharp peaks that would otherwise overwhelm the application layer. Deployed in front of the existing stack with no changes required to downstream services.
.NET
Kubernetes
ArgoCD
Redis
Experience
SP Consult AG, Duisburg
Present2022
Software & DevOps Engineer
3y9m- Core application was a legacy VB.NET Framework monolith — the ongoing effort was to progressively extract functionality into independently deployable services on the Kubernetes platform.
- Designed and built the Kubernetes infrastructure from the ground up (Tanzu → RKE2/Rancher), including two full data-centre migrations — working alongside one other DevOps engineer on a product team of four developers.
- Automated deployments end-to-end with Azure DevOps, ArgoCD, and Renovate.
- Implemented full-stack observability: metrics, logs, and traces via Grafana, OpenTelemetry, Thanos, and OpenSearch — including alert handling and incident response.
- Handled production incidents including a cert-manager certificate renewal failure caught and resolved outside working hours, and an OpenSearch upgrade that failed mid-migration — in the latter case, standing up a secondary OpenSearch instance to keep log capture running while restoring the primary.
- Designed and built a traffic-management service with waiting-room functionality to protect retail endpoints during sudden traffic spikes — a custom reverse-proxy layer that queues and admits requests at configurable rates, keeping backend services stable under load.
- Delivered payment processing and fiscal-compliance integrations (TSE/KassenSichV), a lottery tool, and a complete rewrite of the POS application (ongoing).
- Attended KubeCon to stay current with the Kubernetes and cloud-native ecosystem.
- Built a pipeline that reads changes from the database transaction log and emits them as Kafka events — consumed by multiple services including .NET reporting and a kitchen operator display that surfaces and manages new orders in real time.
Kubernetes
ArgoCD
.NET
Java
Kafka
Azure DevOps
OpenTelemetry
Helm
MSSQL
vSphere
C.Ed. Schulte GmbH Zylinderschlossfabrik, Velbert
20222019
Software Developer
7m- Integrated an e-commerce web shop into the company's new ERP system (AP+).
Apprentice Software Developer
2y6m- Dual-education apprenticeship (IHK) with cross-departmental rotations through software, operations, logistics, and sales.
- Technical curriculum: C++, Java, .NET, MS Access, SQL, networking fundamentals, and virtualisation.
- Built Industrie 4.0 applications — IVR/Hotline (REST), PickByLight (WinForms + REST), AppLauncher (WPF), and licence management (Blazor + REST).
- Set up CI/CD pipelines using Continua CI and FinalBuilder.
.NET
Python
PostgreSQL
vSphere
Skills
Languages
.NET
Java
Python
TypeScript
Bash
Orchestration & GitOps
Kubernetes
Helm
Flux
ArgoCD
Rancher
Aspire
Observability
Prometheus
Grafana
Thanos
OpenTelemetry
OpenSearch
Secrets & Identity
OpenBao
External Secrets Operator
cert-manager
Keycloak
Storage & Data
Garage
CloudNativePG
Redis
Kafka (Strimzi)
MSSQL
CI / CD
Woodpecker CI
Azure DevOps
Renovate
Networking
Cilium
Envoy Gateway
ingress-nginx
Platform & OS
Talos
RKE2
Linux
Docker
Hypervisors
Proxmox
vSphere
Source Control
Git
SVN
Working with AI
AI is part of my toolkit — the same way linters, formatters, and good documentation always were. The surface area is bigger; the accountability isn't.
- I use it to move faster on problems I already have a handle on. Not to skip understanding the ones I don't.
- The architecture and the review are mine — AI is a fast collaborator, not a shortcut.
- I run AI tools inside microVM sandboxes with dedicated, often scoped repository context — better signal in, contained access, better output out.
Contact
Open to full-time senior platform engineering or SRE roles, and available for consulting engagements.