Technical writing3 min read
What a confusion matrix actually tells you
Four numbers decide whether a prediction model is useful, and the two kinds of wrong rarely cost the same.
Machine learning, and the software to actually put it to work.
Data in. Decisions out.
WHAT I DO
01
Data analytics
Dashboards and reporting that answer the questions you actually ask.
02Full stack development
Web applications, internal tools, and the APIs holding them together.
03Machine learning
Forecasting, classification, and models shipped into something usable.
04Game development
Unity gameplay and prototype work, taken on as a separate track.
LATEST WRITING
All writingML NEWS
ARXIVLLMSep 23, 2026
A JEPA Recipe for Tabular Foundation Models
ARXIVLLMSep 23, 2026