Notes from daily engineering work.
Thoughts, lessons and tutorials from real projects – clear, pragmatic, no buzzword bingo.
My Practical Experience with RAG: How Documents Become Reliable AI Answers
How does a PDF become a reliable, source-based AI answer? In this article, I show how I implemented a complete RAG pipeline in Python during my MIT continuing education – from chunking and embeddings to vector search, controlled prompting, and source attribution. This is not about protected course content, but about the technical principles, quality factors, and practical insights behind a robust RAG solution.
What I Learned in Machine Learning: From Dataset to Robust Decision
In my MIT further education in Applied AI & Data Science, I applied Machine Learning from the ground up using practical projects: from clustering, regression, and classification to Random Forest, XGBoost, and time series analysis, all the way to Generative AI and RAG. In this article, I show what I learned about data quality, model selection, evaluation, interpretability, and the journey from a notebook to a robust decision.
JavaScript, TypeScript and the Reality of Complex Software Projects
Personal lessons from automotive projects at BMW – why good software consists not only of working code, but of structure, responsibility, and a team that can handle complexity.
AI in Consulting: Where Language Models Really Help (and Where They Don't)
An honest look at AI integration in consulting projects – beyond the hype.
Backend Development, Spring Boot, and the Reality of Production Cloud Systems
What I've learned from enterprise and automotive projects with Java, Spring Boot, and AWS – why good backends are more than just working APIs, and where configuration details like `spring.jpa.open-in-view` suddenly become critical.
Microservices with Spring Boot: What I Learned at Audi
Practical lessons from several years of backend development in a large, regulated automotive environment.
Quality Assurance in IT – More Than Just Testing
Why quality doesn't just emerge at the end of a project through testing – personal lessons from automotive projects at BMW Crowd Data Collector, transitioning between development and test/validation, and working with DLT logs, Kibana & Co.
Why "It Can't Be Done" Is Rarely True in Software Development
A brief reflection on how a clear mindset and common sense can defuse almost any technical problem.
QA-Enforcer – Quality Assurance as a Connecting Element in Complex Systems
Why a bug is rarely just an error in the code – and how the role of the "QA-Enforcer" connects testing, development, architecture, monitoring, and communication in enterprise systems.
Scrum or Micromanagement? Or simply: Humanity in the Team?
Why "being agile" doesn't depend on frameworks, but on the courage to understand your own team's reality – and to act situationally instead of dogmatically.
Why Every Large Company Needs a Chief Simplicity Officer
Complexity grows quietly – until it becomes a real obstacle. Why it's time for a dedicated role that champions simplicity as a strategic competitive advantage.
MIT Applied AI & Data Science Program – completed 🥳
Six months, 12+ hours a week, from mathematics & statistics to Generative AI: A look back at the MIT program that gave me a structured, end-to-end understanding of AI.
Technical SEO Consulting – An Important Early Career Step
Looking back on my time as a technical SEO consultant at CatBird Seat – why this early position shaped my perspective on client communication, business goals, and technical decisions to this day.
Did a topic spark your interest?
Drop me a line – I enjoy every conversation.