Build the infrastructure your team can actually run.
I'm a DevOps and cloud consultant. I help startups turn fragile, ad-hoc cloud and Kubernetes setups into declarative, observable platforms that people and AI agents can change safely, on a fractional basis: senior judgment and hands-on work without hiring a full-time Head of Infrastructure too early.
DevOps consulting makes sense when the way software reaches production has become a brake: on the team, on the product or on the budget. My job is to understand where the company stands, help the team define a strategy and support them through the implementation until they can run it on their own.
I work alongside small, often overloaded teams that need to grow and become autonomous. As a fractional lead I bring strategy and experienced guidance: I help the team pick the right goals and avoid spending energy on work that does not deliver results.
Design, implementation and operation of Kubernetes, when it is really needed. In the cloud with EKS, GKE and AKS, in the datacenter with OpenShift, at the edge in industrial plants with k3s and MicroK8s. GitOps releases, security, maintenance procedures, costs and, in the end, a team able to run everything on its own.
Design and operation of cloud infrastructure: architecture on AWS and Google Cloud, provider selection, incremental and reversible migrations, costs under control. With one simple principle: keep complexity in check.
I help industrial companies and IoT startups bring Kubernetes to datacenters and industrial plants: lightweight distributions for the edge, automated releases and application architectures that keep working even when the connection drops.
Turning infrastructure that grew in a disorderly way into a self-service, declarative and observable platform: a single, reliable way to create services and ship, which people and AI agents can work on safely.
Design of open source observability systems based on OpenTelemetry for collecting and processing logs, metrics and traces. Incremental adoption of OpenTelemetry to get a monitoring and alerting system integrated with AI agents.
When a team adopts a new technology, a shared skills baseline is often missing. Our training courses are designed to bring the whole team to a common level. They are taught by instructors with real experience running cloud platforms.
I work in parallel for a small number of clients, with continuous responsibility for a specific slice of their technical work. It is neither a one-off consulting session nor project freelancing: it's a responsibility taken over time, on a clearly bounded slice.
A small number, usually two or three.
It depends on the shape. A project can run a few weeks or months. A fractional engagement, such as team mentoring, typically starts at three months and runs for as long as it makes sense. The goal is to no longer be needed.
Yes, I implement when it makes sense, for example in DevOps consulting or a cloud migration. I use agentic code tools to make that practical at this scale. Modern infrastructure (declarative, IaC-first, well-documented) is exactly the kind of codebase where these tools work. The decisions still come from engineering judgment; the implementation time goes down.
I help you define the role, evaluate candidates, and run the transition. That's part of team mentoring, not a side effect.
28 Oct 2026
Do you need Istio?
14 Sep 2026
OpenTelemetry doesn't do alerting. By design: it collects and ships logs, metrics and traces, and stops there . Alerts are left to an observability backend. For alerts in an OpenTelemetry stack, I…