This paper introduces a novel sign language recognition system aimed at eliminating communication challenges faced by individuals with hearing or speech disabilities. The system translates hand gestures into text and audio using a Convolutional Neural Network (CNN) optimized for recognizing 36 static gestures, encompassing the English alphabet (A–Z) and digits (0–9). With a model accuracy of 95%, it demonstrates reliable gesture classification through layered convolutional, pooling, and fully connected operations. Furthermore, it incorporates real-time text-to-speech (TTS) functionality and translates output into regional Indian languages such as Hindi and Marathi. This inclusive feature enhances communication accessibility and supports widespread adoption in multilingual communities.