This report presents the design and implementation of a real-time American Sign Language (ASL) recognition system using computer vision and deep learning techniques. We use MediaPipe for hand landmark detection, OpenCV for image pre-processing and a Convolution Neural Network (CNN) for gesture classification. The system provides a solution to bridge the communication gap between sign language users and non-signers by converting hand gestures into text and speech output in an accessible and affordable way. The project presents a lightweight webcam-based solution that does not require any specialized hardware like sensor gloves or depth cameras. The system uses skeletal hand landmark rendering to improve robustness and accuracy, reducing the effect of background noise and lighting changes. Experiments show that the method has high recognition accuracy and efficient real-time performance, which makes it applicable to practical applications such as educational support, assistive communication, and public service interaction. This work showcases the promise of the integration of artificial intelligence, computer vision and human-computer interaction technologies for the development of scalable and inclusive communication systems. The report also discusses system limitations and future enhancements, including dynamic gesture recognition, sentence-level translation, and deployment on mobile and embedded platforms.