01
Builder
Each strategy is a structured definition: indicators, conditions, and risk. Editable, versionable, and comparable.
MVP · forex and crypto
Oryon brings the strategy builder and backtesting lab together in one place. AI suggests, the visual editor tweaks, and the engine checks.
Optional local execution · MetaTrader 5 via bridge
net return
+12.42%
1,500 candles
win rate
41.7%
max dd
6.81%
sharpe
1.24
synthetic data · reproducible result
Four modules share the same strategy format: what you test, save, and trade is always the same.
01Laboratory
Build your strategy, run a backtest with costs, and check its robustness with walk-forward and sensitivity analysis.
02Library
Each save creates an immutable version. Compare two versions and deploy the one that holds up to the data.
03Trade
Connect MetaTrader 5 via the local bridge and let the engine open and close trades for you, in demo or live with explicit permission.
04Copy trading
Subscribe to public MQL5 signals or mirror a source account with its investor password. Risk limits are yours.
Oryon doesn't promise profits: it measures, executes what you ask, and logs every decision. Always start in demo.
Reproducible
Same data, same result
Transparent
Every cost, visible
No hype
Zero profit promises
01
Each strategy is a structured definition: indicators, conditions, and risk. Editable, versionable, and comparable.
02
AI suggests the definition. You review it in the visual editor and the engine validates it deterministically.
03
Backtesting with costs, equity curve, drawdown, and separate metrics. No magic scores.
start in one minute
Pick a preset, adjust the parameters, and run the backtest. No signup, no card, no promises.
No profit promises. No real execution yet. Only definition, validation, and measurement.
try nowEMA 20/50 cross
Tendencia clásica: la media rápida cruza a la lenta.
net return
-10,38%
max dd
14,56%
win rate
24,0%
trades
25
La muestra tiene 1500 velas y 25 operaciones cerradas. El tamaño es razonable para explorar, no para confiar a ciegas.
Un backtest mide el pasado con costes estimados. Antes de operar en real, pásalo por demo y por walk-forward.
El proveedor de datos no respondió: la demo se ha calculado con datos sintéticos. El resultado sigue siendo reproducible, pero no es mercado real.