Every candle on forex, metals, indices, crypto and Deriv synthetics is analysed by an ensemble of ML models. You get ranked trade setups with entry, stop-loss and targets — each one carrying its measured, out-of-sample win rate.
*Walk-forward win rate on data the models never saw during training, validated models only. Not a guarantee of future results.
A preview of what the engine found on its last run. Members see the full list, exact levels, charts and probabilities.
Everything a discretionary trader checks by eye, computed for every instrument on every run.
Gradient boosting, random forest and logistic regression vote on ~45 engineered features: trend, momentum, volatility, candle anatomy, spikes and session timing.
Engulfing, pin bars, morning/evening stars, soldiers and crows, inside bars and momentum candles are detected and explained on every chart.
H1 and H4 signals are checked against the daily trend. Setups that align with the higher timeframe rank higher.
Entry, volatility-based stop, three targets, expiry time and the expected next-24h range — sized in ATR, not guesswork.
Models are tested only on unseen future data across several folds. A setup is "Validated" only if it beats the baseline after costs, consistently.
Candles with entry, stop, targets and key support / resistance drawn on top. Ranked by an opportunity score from 0 to 100.
Fresh H1, H4 and daily candles are pulled for every enabled instrument — the same markets you see in MT5.
Each historical bar is labelled by what happened next: did price reach the target before the stop?
The ensemble is retrained and tested on later, unseen data. Win rate, expectancy and drawdown are recorded.
Today's candle is scored. Only conviction signals become setups, graded A/B/C and ranked by opportunity score.
No cherry-picking. Each setup is tracked to its take-profit, stop-loss or expiry and added to a public-to-members track record.
No. Markets are uncertain and no model is right every time. The engine shows the probability, the historical out-of-sample win rate and the sample size so you can judge each setup — and it labels setups without a demonstrated edge as "Speculative".
Forex majors and minors, gold and silver, major indices, BTC and ETH, and Deriv synthetic indices (Volatility, Boom/Crash, Step, Jump). Symbol names match what you see in MT5.
Volatility-style indices are generated by random processes, so honest validation usually finds little or no edge there — and the engine will tell you so by marking them Speculative. Boom/Crash and Step indices have structural quirks the models can sometimes pick up.
A: validated edge and aligned with the daily trend. B: validated edge. C: the model sees a setup, but historical validation did not prove an edge — treat it as speculative.
Place the entry, stop and target in MT5 and size the position so the stop-loss risks a small, fixed percentage of your account (1% or less). Setups expire after their horizon — don't chase stale signals.
Create your free account and see today's full list of ranked setups.
Get started free →