Now
Last updated
This is a now page — a snapshot of what has my attention at the moment, rather than a résumé of everything I have ever done. It changes as my focus does.
Building
- GridMind (F1 Predictor) is a race-prediction pipeline powered by ML.NET, utilising SDCA classifiers and Monte Carlo championship simulation. It is built on live Formula 1 data from OpenF1. Currently, I am progressing through the AutoML upgrade path, moving from the hand-tuned SDCA baseline to the AutoML API in ML.NET.
Learning
- RAG (Retrieval-Augmented Generation) and embeddings are integral to GridMind's planned AI analysis layer. This will enable semantic search across past races using embedded fact sheets and provide a hosted service for chat and embeddings in production.
Exploring
- How AI tooling changes the day-to-day shape of engineering work.
- Testing strategies that survive a refactor instead of blocking one.
Writing
- Notes on architecture decisions and the trade-offs behind them.
- Step-by-step guides for the problems that cost me the most time.