Engineering Blog

By engineers, for engineers.

Technical deep-dives, architecture decisions, and lessons from the field.

You Did Everything Right — And Your Supply Chain Got Compromised Anyway

The tool that scanned for threats became one. One green checkmark. One click to merge. And just like that - every secret gone.

August 13, 20266 min read

Valentin Goronjic

Java

Who’s keeping them? – Java Objects, Caching and OutOfMemory errors

How caching tricks Java's garbage collector into keeping dead objects, why that leads to OutOfMemory errors, and how to hunt down loiterers with OQL.

July 31, 20267 min read

Robert C. Kahlert

AI paradigm shift Artificial Intelligence

CapEx vs OpEx: How Every IT Paradigm Shift Moves Your Cost Structure

Every IT paradigm shift moves costs from CapEx to OpEx. Why that isn't automatically cheaper, and how to combine both models sensibly.

July 17, 20268 min read

Damjan Gjurovski

Culture as Code: Building a Team Playbook for Quality Software

Great software isn't built by better tools, but by better teams. Why treating culture like code, designing, reviewing and refactoring it, is what drives quality.

July 3, 20265 min read

Philipp Seifert-Kehrer

AI paradigm shift Artificial Intelligence

IT Paradigm Shift: How Risk Management Decides the Success of Your Software Development

Every IT paradigm shift moves your risk profile. Which mode fits your company - Agile, Microservices, Cloud, or AI?

June 18, 20266 min read

Damjan Gjurovski

Research Observability Green Native

Green Native Interactive Attribution Model

End-to-end carbon attribution turns IT infrastructure into one number: carbon per transaction for the business, and a drill-down path for engineers. Our Green Native approach builds on metrics cloud-native tools already collect, so measuring adds almost no overhead.

June 3, 202612 min read

Paul Weißenbach Valentino Zablocki

Artificial Intelligence Insights Modern Workflows

How Our Marketing Team Started Writing Pull Requests

He went in skeptical. He came out convinced. The story of a website rebuild where non-developers shipped most of the code.

May 22, 20265 min read

Valentin Goronjic

Latenzkurven p50, p95 und p99 für /api/orders über 24 h — p99 mit deutlich stärkeren Spitzen bis 300 ms während der Geschäftszeiten.

Observability Prometheus Grafana

Better Grafana Dashboards: Precise Prometheus Metrics Over Visual Illusion

Monitoring tools are indispensable to modern software development and cloud infrastructure. The pairing of Grafana as a visualization platform and Prometheus as a metrics backend is one of the most widely deployed in the field.

May 7, 20264 min read

Leo Bowen Wang Robert C. Kahlert Tobias de Vries

Artificial Intelligence AI paradigm shift

Paradigm Shifts in IT: AI Follows the Same Pattern as Agile, Microservices, and Cloud

The AI Engineering paradigm follows the same pattern as Agile, Microservices, and Cloud. What CTOs need to take away for their strategy.

April 30, 20266 min read

Damjan Gjurovski

Research Insights

A Mid-Term Review After 12 Months of Building a Research Department at Posedio

About a year ago, Paul Weißenbach and Stefan Dumss began driving the build-out of the development department at Posedio. Time for an initial mid-term review: which milestones have been reached, which research projects have been successfully completed – and what is the team currently working on?

April 23, 20263 min read

Paul Weißenbach

Artificial Intelligence

AI-Assisted Development: Why Human Oversight Still Defines Engineering Excellence

AI tools deliver real productivity gains in coding, but the real challenges now lie in design, scalability, and safe operation. When AI autonomy grows faster than human oversight, real risks emerge. At Posedio, we combine AI-driven efficiency with deliberate guardrails - from design-first principles to meaningful code reviews, observability, and security as a continuous practice. Because in the age of autonomous systems, what counts isn't output, but ownership.

April 23, 20266 min read

Fränk Lotzkes Manuel Eiweck Valentin Goronjic

Infrastructure as Code Artificial Intelligence

Infrastructure as Code: Do We Still Need IaC in the Age of AI Agents?

AI Agents können Infrastruktur in Sekunden provisionieren. Macht dies Infrastructure as Code damit überflüssig? Nein - im Gegenteil: Ohne deklarative Guardrails werden manuelle Fehler einfach zu schnelleren, automatisierten Fehlern mit größerem Blast Radius. IaC wird zur Governance-Schicht für autonome Agents. Wir zeigen, warum Policy-as-Code und ein abgestuftes Trust-Level-Framework jetzt entscheidend sind – und welche Rolle NIS-2 und ISO 27001 dabei spielen.

April 23, 20264 min read

Philipp Seifert-Kehrer Severin Leuchtenmüller Stefan Dumss

Engineering Blog

You Did Everything Right — And Your Supply Chain Got Compromised Anyway

More from the Blog

Who’s keeping them? – Java Objects, Caching and OutOfMemory errors

CapEx vs OpEx: How Every IT Paradigm Shift Moves Your Cost Structure

Culture as Code: Building a Team Playbook for Quality Software

IT Paradigm Shift: How Risk Management Decides the Success of Your Software Development

Green Native Interactive Attribution Model

How Our Marketing Team Started Writing Pull Requests

Better Grafana Dashboards: Precise Prometheus Metrics Over Visual Illusion

Paradigm Shifts in IT: AI Follows the Same Pattern as Agile, Microservices, and Cloud

A Mid-Term Review After 12 Months of Building a Research Department at Posedio

AI-Assisted Development: Why Human Oversight Still Defines Engineering Excellence

Infrastructure as Code: Do We Still Need IaC in the Age of AI Agents?