Hands-on engineering.
Ownership through production.

How I work

I work directly with CTOs, engineering leads and platform teams, remotely and in English. You work with the same person from architecture and implementation through deployment, operations and handover.

A focused review, a defined delivery project or ongoing engineering support can all be useful starting points. We agree the scope, deliverables and responsibilities around the problem your team needs to solve.

Where I fit

My work connects application code with the platform it runs on: backend services and AI integrations, containerized delivery, cloud infrastructure, security and production operations.

I have particular experience in healthcare environments, Kubernetes, private networks, restricted egress, and hybrid or self-hosted infrastructure. I help teams make technical trade-offs explicit and prepare to own the result.

Previous work

Satelia · 2025–2026: CI migration from GitLab to GitHub Actions, Terraform adoption, container security, release operations and production on-call. Docker Scout scores improved from F to A/B.

Owkin · 2023–2024: Go orchestration work for Substra, Kubernetes Pod Security Policy migration to Pod Security Admission Restricted, and controlled network access for Jupyter workflows.

A wider engineering background

At Heywear, I worked across backend, mobile, web and production infrastructure as a principal engineer. At HARMAN / Samsung Bixby, my intent-processing approach was adopted by eleven international localization teams.

That background helps me work across the boundaries between a product feature, its delivery pipeline and the system that runs it.

Download my CV (PDF) ↓

Containers and community

I am a Docker Captain. My past projects include buildgraph, an open-source CLI I built to understand BuildKit execution, cache behavior and build regressions.

I also write about platform engineering and the operational questions behind AI tooling, including running MCP in production.

Building for AI developers

I created Remainy, a macOS menu bar utility for checking Codex usage and reset times. It is an independent project, unaffiliated with OpenAI.

My focus is practical: useful AI products and developer tools, backed by infrastructure that teams can understand, operate and improve.