Quant

Evidence over advice — all the way down.

Author strategies in plain language, prove them with walk-forward validation and false-discovery-rate correction, and run a pick funnel that logs a reason for every name it dismisses. The rigor of a quant desk, without the desk.

Join the waitlist → Drive it from your editor

Author signals in plain English

Describe a strategy the way you'd explain it to a colleague. The platform writes the Python, runs it on one symbol or the whole universe, and shows you the signal — then you iterate by talking, not debugging.

Natural language → score, strategy, indicator, or custom-metric code
Run on a watchlist or the full universe
Deploy as a recurring subscription when it earns its place

Genie chat authoring a trading signal alongside the live dashboard

The Strategy Funnel — with receipts

A multi-stage pipeline takes the universe down to a ranked shortlist: pre-flight freshness, sector rotation, convergence, edge check, an AI ensemble, validation, tiering. Every run ends with an audit table and a dismissal log — every name that got cut shows the gate it failed. No black box.

Universe → convergence → edge → AI ensemble → validation → tier
Audit table: how many survived each stage
Dismissal log: the measurable reason every cut name was dropped

Strategy Funnel dashboard — universe-wide picks ranked by conviction

Walk-forward validation, not curve-fits

Every strategy carries a trust label derived from a real train/test split — validated, caution, likely-overfit, insufficient-data. We measure the in-sample vs out-of-sample Sharpe, flag the overfit ratio, surface decay, and refuse to promote anything that fails the gate.

Per-symbol IS/OOS split with an overfit ratio
Universe-wide cross-sectional pooling for low-frequency strategies
Decay metrics — surfaces edges that have stopped working

Score-trend chart showing a strategy's conviction evolving over time

Statistical rigor built in

Run enough backtests and something will look brilliant by luck. The platform applies false-discovery-rate correction across multiple tests, validates across volatility regimes, and checks signal correlations so you're not fooled by three indicators that are really the same one.

FDR (Benjamini-Hochberg) correction for multiple testing
Regime-stratified validation across VIX regimes
Correlation matrix to catch circular confirmation

Strategy anatomy view with backtest metrics and validation detail

Convergence — make the strategies agree

One signal is a hint. Stack up to twelve conditions with AND logic, route by cohort, blackout earnings, and the universe collapses to the handful of names where your strategies actually converge. That's where conviction comes from.

Up to 12 co-conditions, ANDed together
Cohort routing and earnings-blackout filters
A vote count per symbol — how many strategies say BUY

Convergence picks dashboard with conviction bars and signal markers

Decision support, never directives

We hand you the evidence. You make the call.

TraderGenie is a decision-support tool, not investment advice. Past performance isn't future results, and every label on the platform is there so you can judge for yourself.

Join the waitlist →