Engineering range
Three things we built because they were hard.
These are our own research projects, not client work. They cover the parts of engineering that are difficult to fake: real-time signal processing, computer vision, and systems programming. If your build needs any of that, this is the evidence.
Research prototype
Elliott Wave trading engine
An automated market-structure engine that identifies Elliott Wave patterns across price series and generates wave counts and signal annotations in real time. It is not a trading product, and none of its analysis is investment advice.
What it proves: Streaming market data, stateful pattern detection, and sub-second annotation under load.
Research prototype
Sign-language recognition research
Computer-vision research recognizing sign-language gestures from video and translating them into text.
What it proves: Real-time gesture recognition, sequence modeling, and applied machine-learning depth.
Research prototype
Rust CLI agent
A command-line AI agent written in Rust: a fast, low-overhead agent loop running without a managed runtime underneath it.
What it proves: Systems-level engineering and an agent loop with no runtime between it and the machine.
None of this is for sale; the catalog is what is for sale. It is here because a buyer evaluating an engineering team deserves to see something harder than a marketing site.