pdtbench — an agentic trading benchmark with a control group
A language model makes money on a historical market window. Is that skill, or does it remember the chart? Every real window gets a synthetic twin, so you can tell the difference.
selected work
05Machine learning systems built against a constraint — a latency budget, a user who cannot see the screen, or a benchmark designed to catch me fooling myself. Each one has a full write-up.
Penny — voice-first banking for blind and low-vision customers
Banking for blind and low-vision customers, where voice and sound are the primary channels and the screen is secondary. Nothing executes until Penny says what will happen and you confirm it.
Byte the Evidence — an adversarial negotiation trainer for junior M&A lawyers
An Ace Attorney-style trainer where junior M&A lawyers negotiate against AI opposing counsel — catching the errors it plants and drawing out the facts it hides, from LOI through to signing.
λGPT — a natural-language REPL built from parser combinators, not an LLM
A conversational interface with no model behind it. Every intent is a parser, every answer is derived, and nothing is generated.
COVID-19 mortality risk — logistic regression from scratch, and a negative result
Logistic regression written from scratch in NumPy — sigmoid, cross-entropy and hand-derived gradients. It scored 75.82% accuracy against a 75.63% baseline, which is to say it learned almost nothing.