Problem framing
How the shape of the question determines what technical work is worth doing in the first place.
AI, SOFTWARE, AND SYSTEMS THROUGH A PRACTICAL LENS
Notes on machine learning, software systems, and the point where technical judgment has to meet workflows, planning, and business context.
Current Exploration
Learning AI backwards, from LLM curiosity into the ML foundations behind modern systems, then revisiting the foundations in the proper order.
Recurring themes
Much of the useful work happens before modeling starts: framing the problem, defining the data, and understanding what people need to trust.
How the shape of the question determines what technical work is worth doing in the first place.
How the picture changes once the data is defined carefully and the measures reflect real use.
How forecasts become useful when they are tied to timing, constraints, and actual planning decisions.
Whether a system fits the workflow around it and can keep being used after the initial enthusiasm fades.
Writing
A personal note on learning AI backwards, from LLM curiosity into the ML foundations behind modern systems.
Read noteA short note on Euler's number and why e keeps appearing in growth, compounding, and continuous change.
Read noteGet in touch
If something here connects with work, planning, or a question you are thinking through, a note is welcome. I read every message myself.
A few lines of context are fine.
Go to contact