A sixteen-pair forex signal engine built on Lorenz manifolds, persistent homology, and Aristotelian decision logic. Live, running, trading.
16 vertices · 16 pairs · 16 founding seats
Forex markets look random. They aren't — they are almost random, perturbed by slow geometric structure that shows up in the correlations between pairs over minutes and hours. Our engine reads that structure in four ways at once: Lorenz-attractor wings for regime detection, persistent homology for topological breaks, differential curvature for local torsion, and a rule system modeled on Aristotelian categorical logic for entering and exiting positions.
No promises of returns, no backtest cherry-picking, no mysticism. Just an honest research project, running in public, producing real trades against a real broker account. You can watch it think further down this page.
Founding members fund the core research and get the dashboard that no one else sees: live positions, every signal component as it updates, the full decision trace with the reasoning for each entry and exit, and the downloadable P&L archive across all sixteen pairs at millisecond resolution.
You also get trade alerts the moment the engine enters a position (email, optional SMS), the backtest notebooks we run internally, and a direct line to the research team via monthly office hours. When we publish — whitepapers, public talks, open-source modules — you get it first.
Running this is work: paying for the data feeds, paying for the compute, babysitting the engine, writing the math, fighting the broker when it misbehaves. A founding membership is how you fund that work without having to do any of it yourself. I go to the market; you watch; we share what we find.
Every dollar is accounted for — sixty percent of any trading profit is routed back, by design, into what the system calls the generosity tensor: a fixed redistribution across members, pharma research, and open-source algorithmic work. You are not buying a product. You are seeding a research program.