Notes on data, AI, IT
and security
No marketing fog. The way I think about real problems with founders and managers.
IT system resilience when conditions shift fast
How a manager should think about IT infrastructure resilience when the external environment changes quickly and unpredictably.
Data mesh is about ownership, not about the platform
Breaking down the data mesh concept without the hype - why it is an organisational model first and a technical stack second.
Log4Shell: the management lessons from the incident
Breaking down the Log4Shell vulnerability as a management lesson - about hidden dependencies, response speed, and invisible risk.
GPT-3 in the API: what a founder should do with it
OpenAI opened GPT-3 access through its API. A clear-headed look at what changes for business and where to slow down.
Log4Shell: if you do not know your dependencies, you do not know your attack surface
The Log4Shell vulnerability showed that most companies have no idea which libraries are running inside their systems.
Data ownership: who signs off on the number
In most companies data exists but no one is responsible for its quality. I look at what data ownership actually means in practice.
Zero trust: what it actually means and when it is worth the investment
Zero trust has become one of the biggest buzzwords in security. I break down what is behind it and who it is actually relevant for.
When to split a monolith: the questions matter more than the hype
Microservices are a popular answer to the scaling question. But the right question is not 'split or not' - it is 'why and when'.
Event-driven architecture: what managers need to know before committing
Events and message queues solve real coordination problems between services. They also introduce complexity that is easy to underestimate from a project plan.
The gap between experiment and production: why ML models never reach work
Most ML projects show good results in experiments and perform poorly in production. I look at why this happens.
When a good model goes bad: drift, detection, and business cost
A model that passed every test at launch can quietly degrade over months. Understanding why helps you decide how much monitoring is worth the investment.
Real-time analytics: when it works and when it is expensive theatre
Streaming data and real-time dashboards have become a fashionable requirement. I look at when this actually solves a real problem.