Single Ticker
Multi-factor BUY/HOLD/SELL output with confidence, vote consensus, targets, and stop-loss context.
AI Stock Signal Research Platform
Analyze single tickers, scan sectors, backtest outcomes, and explore options ideas in one integrated workflow. Built for transparent research and decision support, not trade execution.
Checking live engine...
Current RL Training Stage
Loading RL status...
0 / 500 steps to next milestone
RL Training Stages
| Steps | Stage | Interpretation |
|---|---|---|
| 0-500 | Infant | Mostly exploratory behavior |
| 500-5,000 | Learning | Early pattern formation |
| 5,000-25,000 | Developing | Improving but still unstable by regime |
| 25,000-100,000 | Intermediate | Stronger pattern recognition and consistency |
| 100,000+ | Experienced | High training volume, still requires ongoing validation |
Universe Scan
Rank broad symbol sets by sector and horizon to surface top directional opportunities.
Backtest and KPI Reports
Validate behavior with walk-forward summaries and performance leaderboards over rolling windows.
How Signals Are Actually Generated
Q Signals combines 23 analysis modules into one composite score, then applies a two-gate decision system before issuing BUY, HOLD, or SELL.
Gate 1: Composite Threshold
BUY requires composite score >= +0.10. SELL requires composite score <= -0.10. Scores between these thresholds lean neutral unless other evidence is strong.
Gate 2: Module Agreement
At least 60% of modules must agree on direction. This prevents weak one-factor spikes from becoming trade signals.
Confidence and RL Agreement
Confidence increases as score clears threshold with stronger agreement. RL agreement indicates whether the DQN agent independently reached the same action.
23-Module Coverage (Transparent Breakdown)
Modules include technical structure (Fibonacci, Elliott, swing, volume, multi-timeframe, trend, volatility, moving averages), macro/regime context (VIX, sector rotation, correlation, economics, Fed), flow/positioning (insider, institutional, congress, competitor, supply chain), sentiment (news/social), analyst inputs, and an RL module.
Targets, Stop, and Hold Window
Every directional signal includes conservative/moderate/aggressive targets, a stop-loss, and a specific hold duration based on ATR and selected horizon. Outcome evaluation checks whether target or stop was hit first by the deadline.
RL Training Stages (Honest Labels)
Infant (0-500), Learning (500-5K), Developing (5K-25K), Intermediate (25K-100K), Experienced (100K+). Stages reflect training volume, not guaranteed performance.
Not investment advice. This is a research system with non-stationary market risk. Use as decision support, not blind execution.
Recent Live Signals
Pulled from your existing Supabase-backed API in near real time.
Market Insight Pages
SEO landing pages that help users discover and test Q Signals workflows.
Daily Financial News
Fresh headlines for market context and ticker-specific research workflows.
Education Center
Understand indicators, confidence, risk overlays, and portfolio controls before tuning settings.