Research area
Notes from the Ferro engines — feed normalization, the matching core, options analytics, and the wavelet signal layer. Mostly systems work: latency, memory layout, and the numerical methods underneath.
5 posts
Time-series ML fails quietly when the feature layer ignores scale. A price series, volatility series, or sensor stream may look like one signal, but its behavior is usually a mixture of fast noise, medium-horizon stru…
A Rust options pricing engine usually gets judged by surface area first: how many models, how many Greeks, how fast the implied-volatility solver runs. That is necessary, but it is not enough. The harder question is w…
Market-data infrastructure becomes fragile when every downstream component learns a different venue dialect. One feed calls it a trade condition, another calls it a flag. One stream uses exchange timestamps, another c…
Price-time priority sounds simple until the implementation has to survive real state transitions. Orders arrive, cancel, partially fill, cross multiple price levels, and interact with timestamps, client identifiers, a…
An implied-volatility surface is a deceptively small object: a function from log-moneyness and maturity to a single number. The trouble is that not every such function corresponds to a market. A surface that prices a …