Any thriving and competitive economy depends heavily on the stock market. By making shares of a company publicly available, it aids in its financial growth. Additionally, it enables people to invest in those businesses and earn from doing so. However, due to the volatility of the stock market, stock trading entails a certain risk. This study aims to minimize this risk by providing a solution for stock price prediction. The solution is based on predicting the stock price using a combination of technical and sentimental indicators. The used technical indicators are SMA, EMA, MACD, RSI, and Momentum. To measure their influence on the stock price, sentiment analysis is applied to the stock-related tweets. The BERT model is used for sentiment analysis with a classification accuracy of 82.88\%. The sentimental indicators are derived from the sentiment score, and they are: Averaged Weighted Sentiment, Sentiment Moving Average, Tweets Volume, and Tweets Moving Average. The technical and the sentimental data are then fed to the machine learning models i.e. LSTM, and GRU to predict the stock price. During the experiment, the RMSE is reduced from 13.83 to 0.33. The first RMSE results from using LSTM with only technical indicators. The second RMSE results from using GRU with a set of technical and sentimental indicators. These numbers show how the sentimental analysis contributes to better stock price prediction. This study also demonstrates the superiority of GRU over LSTM in the field of stock market prediction.