A machine-learning project that predicts stock price movements by combining historical data with real-time sentiment. Uses financial APIs, FinBERT for sentiment extraction, and a Hybrid LSTM model achieving 73.2% accuracy, showing sentiment strongly improves prediction.