Notes on data, AI, IT
and security
No marketing fog. The way I think about real problems with founders and managers.
AlphaGo and the shift in what we expect from AI
AlphaGo's win over Lee Sedol is not just a technical result. It is the moment when the AI conversation stops being only about recognition.
Technical debt: it is an ownership question, not a metaphor
Technical debt gets discussed as a technical concept. For leadership it is first and foremost a question of who makes decisions and lives with the consequences.
Why structuring data must come before any ML model
Before the conversation reaches algorithm selection, you need to establish whether there is data worth learning from. I walk through that step in detail.
Breach notification delay: the management risk that gets underestimated
Companies discover breaches months after the event. I look at why this is a leadership problem, not just a security team problem.
Microservices: the management cost nobody talks about
Moving to microservices is not purely a technical decision. It introduces real operational load on the organisation that rarely appears in the pitch.
TensorFlow goes open source: what changes for non-researchers
Google opened TensorFlow in November 2015. I look at what this means for companies that are not in the business of academic research.
Energy grids and ICS security: the control system is no longer separate from cyber risk
Why attacks on industrial control systems have become a reality for the energy sector, and how managers need to rethink security for OT infrastructure.
Lift-and-shift cloud migration: what is hidden in the bill
Why moving applications to the cloud without rearchitecting often costs more than expected, and what to examine before starting a migration.
Industrial robots: ROI depends on the process, not the robot
Why most mistakes in manufacturing automation happen before the robot is purchased - and how a manager can evaluate real returns.
Streaming data: when operational decisions cannot wait for a batch
When a business needs streaming instead of batch processing, and what needs to be decided before adopting Kafka or similar tools.
TensorFlow and the shift of machine learning from research to engineering
What Google's open release of TensorFlow changes for companies: pipeline, reproducibility, and deployment become the central question, not algorithms.
NLP text classification as a practical enterprise baseline
Before the deep learning wave reshaped NLP, classical text classification already solved real problems. What it does well, where it stops, and how to start.