Institutional
Methodology.
StockSage uses FinBERT financial NLP and multi-factor technical analysis to generate trade signals for Indian equities. All signals are backtested — confidence percentages reflect historical win rates, not predictions.
Data Ingestion
End-of-day data for 100 NSE equities and 3 major indices, stored in Neon PostgreSQL via Drizzle ORM.
Signal Engine
FinBERT NLP combined with RSI and volume analysis to generate multi-factor signals. Confidence values are backtested historical win rates — educational indicators, not investment advice.
Signal Methodology
Each signal requires 3+ concurring factors before firing — no signal triggers on a single indicator alone.
BUY — Oversold Bounce
RSI < 35 + Bullish Sentiment + Normal Volume
BUY — Momentum Spike
Volume > 1.5x + RSI 55-70 + Bullish Sentiment
SELL — Mean Reversion
RSI > 70 + Bearish Sentiment + Normal Volume
SELL — Bearish Dump
Volume > 1.5x + RSI < 40 + Bearish Sentiment
StockSage is not SEBI registered. All signals are educational indicators, not investment advice.