Speech, our natural communication tool, conveys not just words, but emotions too. Speech Emotion Recognition (SER) aims to decode these hidden feelings, bridging the gap between humans and machines. While annotating emotions and capturing their subjectivity pose challenges, SER unlocks immense potential. Our work proposes an ANN-based system to extract emotions from speech, focusing on crucial acoustic features. Leveraging existing datasets and models, we trained our ANN to recognize seven emotions: anger, disgust, fear, happiness, neutral, sadness, and surprise. The proposed model achieved remarkable accuracy, reaching 100% during training and 99% during validation. This work contributes to advancing SER technology, paving the way for more empathetic human-computer interactions.