Content & Talks

What does it take to migrate the deployment of 30+ Kubernetes clusters from a patchwork of CDK and Sceptre to Terraform without breaking production? In this comprehensive post, I share our team’s journey at Adevinta, covering the automated migration strategies, custom tooling decisions, and hard-learned lessons about rollbacks and local development workflows.

From simple Go-based import automation to sophisticated CI pipelines that prevented state conflicts, we built confidence through iterative waves—starting with low-risk IAM roles and progressing to critical VPCs and network routes. The post reveals why custom tooling sometimes beats off-the-shelf solutions and how hands-on learning scaled better than documentation for complex technical migrations.

How do you navigate cryptographic export compliance when publishing a Flutter app to the Apple App Store? This post recounts my journey through the maze of legal requirements, technical investigation, and the surprising complexity behind a seemingly simple HTTPS connection. I share the confusion many developers face, the technical deep-dive into how Flutter handles encryption, and the critical importance of making the right compliance declaration to avoid legal and account risks.

Managing Helm chart sources can make or break your Kubernetes deployment workflow. In this post, I compare two main strategies—centralized chart repositories versus collocating charts with application code—highlighting the trade-offs, organizational fit, and practical implications of each. You’ll also get a primer on GitOps and how it changes the deployment landscape.

If you’re looking to streamline your Kubernetes deployments or want to understand the real-world impact of Helm chart management, this summary will help you decide which approach fits your team.

How do you enable thousands of engineers to seamlessly use ARM nodes in Kubernetes without manual intervention? In this article, together with my colleague Miguel Bernabeu, we detail the technical and organizational challenges of introducing ARM architecture into a large-scale, multi-tenant platform. Discover how we leveraged mutating webhooks, optimized multi-arch builds, and open-sourced our solution to make ARM adoption transparent and efficient for over 4,000 developers.

If you’re interested in platform engineering, Kubernetes internals, or cost optimization at scale, this summary will guide you to the full technical journey and open-source release.

What happens when a Kubernetes autoscaler misconfiguration spirals out of control? In this post, together with my colleagues Fabian Selles and Tanat Lokejaroenlarb, we share our experience with a real-world incident story: a minor alert led to a cluster-wide crisis, relentless pod evictions, and a deep debugging journey. Learn how we traced the root cause to a hidden rate limit, the lessons we learned about VPA internals, and how open-source code access saved the day.

In this webinar I share our experience with my colleague Alfredo Lopez, in Adevinta, a leading European online classifieds company, our experience implementing large-scale observability. The talk covers:

• Challenges faced: Adevinta describes the technical and organizational hurdles encountered when adopting a comprehensive observability strategy across multiple teams and platforms.
• Achievements: The benefits highlighted include improved incident detection and resolution, unified system visibility, and the ability to correlate metrics, logs, and traces for more effective analysis.
• Lessons learned: The team shares key takeaways about the importance of cross-team collaboration, standardizing tools and processes, and adapting organizational culture to fully leverage observability.
• Business impact: Thanks to observability, Adevinta has optimized operations, reduced incident response times, and improved the end-user experience.

This talk is a valuable reference for organizations looking to modernize their monitoring and management of complex systems, showcasing both obstacles and best practices for successful observability adoption.