This document describes a cryptocurrency price prediction system called AlgoB that uses an LSTM neural network model. The system was developed by four students to predict cryptocurrency prices with high accuracy. It takes historical price data as input and can predict future prices. The system uses libraries like NumPy, Pandas, TensorFlow and Matplotlib. It achieves 80% prediction accuracy, outperforming regression and tree models. The LSTM model is trained on price data and evaluates predictions against real prices. This helps traders understand market movements and identify good times to buy and sell cryptocurrencies.