Notes on data, AI, IT
and security
No marketing fog. The way I think about real problems with founders and managers.
AI does not fix bad data
A short note on why an AI rollout in a company starts not with the model, but with the quality of the data underneath.
What data engineering is, and why business needs it before AI
Why companies have to gather and structure data before talking about models and agents.
The next evolution of Agents SDK: long tasks, sandbox, and a production-ready agent
Tools for building AI agents are maturing. What this means for companies thinking about real deployments rather than demos.
Excel in a company is not a shame, it is a symptom of growth
Why a company being held together by Excel is not embarrassing, but a signal of which processes have outgrown their tools.
How to tell when your vendor is making the project more complicated than it needs to be
Signals that an executive can pick up early, before the project's problems become impossible to ignore.
Shadow AI: how the new shadow IT is becoming a security problem
Employees are using AI tools without IT department oversight. The pattern is familiar - but the risks are different from classic shadow IT.
Data, IT, and security cannot be separated
Why splitting these three areas across different teams turns any technology project into a quiet source of hidden risk.
How I look at a new technology project
The set of questions I run any new request through, from an AI assistant to industrial analytics.
Long context in LLMs: what it changes for business tasks in 2026
Modern models support context windows of hundreds of thousands of tokens. What this practically changes for companies and where the real limits are.
Simple architecture often beats trendy
How the urge to use the right stack and draw beautiful diagrams quietly breaks projects that could have been running calmly for years.
Internal API governance: why you need it when you have more than three teams
When a company grows to several product teams, internal integrations start creating problems. Why this happens and how to work with it.
Streaming data architecture: when it actually changes an operational decision
Streaming data is a popular topic. But for most business tasks, the more useful question is when it is genuinely needed versus when it is overengineering.