Due to the altering user behavior, the volume of data published by users on numerous social networking sites, including Instagram, Twitter, Snapchat, etc., has significantly increased. As a result, an enormous volume of textual, audio, and video posts-reflecting millions of users' ideas and opinions-are produced daily. Even if some posts are brief and don't matter, they provide insightful personal viewpoints on various subjects. This article employs Twitter as the analytic platform, where posts are known as tweets, focusing on identifying the underlying emotions behind such posts. This study investigates techniques for preparing and extracting Twitter data using Python-TensorFlow. Additionally, this study incorporates Recurrent Neural Network with Bi-Long Short-Term Memory (RNN (Bi-LSTM)) into the code, improving the model's performance. This research aims to effectively extract sentiment from tweets by training and evaluating the data against a classifier.