Article’s

The exponential growth of retail investors in Indian equity markets has created a critical need for accessible, datadriven decision support systems. This paper presents a novel hybrid machine learning platform that combines Random Forest classification, technical analysis, and sentiment analysis to generate actionable trading signals for National Stock Exchange (NSE) listed securities. The system addresses the limitations of pure machine learning approaches, which are prone to overfitting in financial time-series prediction, by integrating domain-specific technical indicators and news sentiment into a weighted ensemble model. The proposed methodology employs Random Forest classifiers trained on 25 technical indicators including RSI, MACD, Bollinger Bands, Stochastic Oscillator, and volume-based metrics, calculated from five years of historical OHLCV data for 15 NSE large-cap stocks. A hybrid signal generation algorithm combines ML predictions (40% weight), technical analysis scores (40% weight), and sentiment analysis (20% weight) to produce BUY/SELL/HOLD recommendations with confidence levels. The system achieves 62.3% average prediction accuracy on test data, outperforming pure ML models (58.1%) and pure technical analysis (54.7%) by 4–8 percentage points. The full-stack implementation integrates the ML pipeline into a production-ready web application featuring paper trading simulation with 100,000 virtual capital, real-time portfolio analytics including Sharpe ratio and maximum drawdown calculations, interactive candlestick charts with technical indicator overlays, and community-driven features for collaborative learning. The platform demonstrates practical deployment of machine learning models in financial applications while maintaining sub-500 ms inference latency for real-time predictions. Experimental results on historical data (2020–2025) demonstrate that portfolios following the hybrid signals achieve an average Sharpe ratio of 1.47, significantly higher than random trading (0.23) and buy-and-hold strategies (0.89). The system successfully handles 100+ concurrent users with 95% of API requests completing within 2 seconds, validating its scalability for educational and research applications. This work contributes a novel hybrid architecture for financial prediction, demonstrates end-to-end ML deployment patterns, and provides an opensource educational platform addressing the accessibility gap in Indian financial technology.