Give it a symbol and a timeframe. AQ Genesis runs the whole research pipeline — search, backtest, validation, survival — and returns a basket of strategies ranked by confidence, not by the prettiest curve.
Most strategy generators optimize for the highest historical return — which is exactly how you manufacture overfitting. Genesis inverts the priority: every candidate must survive the validators (walk-forward, Monte Carlo, out-of-sample, PBO) before its return even matters.
What you get is not "the best backtest" but a basket of survivors, each with its evidence attached.
Genesis doesn't reinvent anything: it drives the same Strategy Lab, robustness suite and portfolio tools you use by hand — the machinery you can inspect, configure and re-run yourself. Every strategy it proposes is fully open: rules readable, parameters visible, exportable.
Exhaustive, random or genetic exploration of entries and exits over your indicator vocabulary.
CPU + GPU engine with realistic costs, slippage and every exit mode of the Lab.
Walk-forward, Monte Carlo bands, out-of-sample and overfitting probability — the gauntlet.
Opposite-signal exits, time stops and trailing modes explored — because the best entry depends on the exit.
A set of low-correlation survivors, ready for the Portfolio Builder and live health monitoring.
Every rule in readable strategy language, exportable to TradeStation EasyLanguage and MQL5.