This paper addresses the crucial goal of improving communication between the deaf and mute community and the broader public. We present the development of a sign language translator that effectively converts hand gestures into text, encompassing alphabets and digits, thereby facilitating comprehension of the conveyed messages. Our approach incorporates an intelligent hand gesture recognition system powered by Convolutional Neural Networks (CNNs) and utilizes a comprehensive dataset comprising 44 distinct gestures, including letters and numbers. The proposed model plays a pivotal role in preprocessing input images by employing a threshold mechanism to eliminate noise and enhance image smoothness. To ensure accuracy and continuity in the regions of interest, we implement a region filling technique. TensorFlow serves as the backend for our CNN-based Keras model, which undergoes rigorous training on the collected dataset. After the training phase, the model is rigorously tested to validate its performance. Upon successful testing, users can express their gestures, and the system displays the corresponding text representation while simultaneously converting it into speech. This groundbreaking work contributes significantly to bridging the communication gap and fostering inclusivity within our society.