System Architecture
Under the hood of not-financial-advice — a 3-engine, AI-powered money maker.
System Overview
The architecture uses scheduled crons to orchestrate data across market intelligence providers, filter through a 3-engine conviction algorithm, and enforce a strict 3:1 Risk/Reward gate via Gemini Flash before persisting any recommendation.
Execution Pipeline
The scanner executes 3x daily. Mock/rate-limited responses are automatically discarded. Buy signals trigger $1,000 paper trades. A separate validation cron closes positions when TP, SL, or a 14-day expiry is hit.
3:1 Risk/Reward Gate
Gemini Flash is explicitly banned from recommending "Buy" unless the profit target is ≥ 3× the stop-loss risk. If the math doesn't pass, the AI must output "Hold".
Paper Trade Lifecycle
Every Buy signal auto-opens a $1,000 simulated position. The validation cron closes it when the price hits the AI's exact take-profit or stop-loss target, or after 14 days.
API Call Budget
| Stage | Calls/Scan | Daily (×3) |
|---|---|---|
| Discovery (Top Gainers) | 1 | 3 |
| Deep Dive (per candidate) | 6 | 18 |
| Validation (price checks) | 10 | 10 |
| Total | 17 | 31 |
Mock Mode & DB Protection
When API limits are hit, mock data is generated for UI display but is explicitly discarded from the database. DB pollution guards prevent fake data from corrupting historical analytics.
The Gainer Fade Thesis
Empirical observation: stocks that gap up significantly on the daily Top Gainers list tend to fade sharply within 1-3 days. The larger the gap, the higher the probability of mean-reversion.
Low Fade Risk
3-8%
Moderate gap, sustainable momentum
Medium Fade Risk
8-15%
Overextended, watch for reversal
High Fade Risk
15%+
Extreme gap, likely to retrace hard
The Scanner UI now highlights the "Fade Risk" level on each candidate card. High-gap candidates that the AI rejects as "Hold" are potential short/fade opportunities.