Markus' Blog

Posts

  • Why Industrial Machine Learning is Hard

    Industrial machine learning has some particular challenges such as scarce labels, skewed data, costly exploration and poor observability. Not all is lost, though: there are a number of techniques to tackle them.

  • Machine Learning: What Could Go Wrong?

    A field guide to how machine learning models fail in practice — from problem-definition exploits to distribution shift — and how to guard against each.

  • How Much Data Do You Need?

    Data only has value in combination with a problem — why the amount of data you need depends on model complexity, not on collecting as much as possible.

  • Agentic Software Development: Take the Director's Chair

    Why 'coding agent' undersells the job, and what it means to direct rather than micromanage AI-assisted development.