In the domain of assistive technology, advancements in artificial intelligence have been able to revolutionize accessibility for individuals with speech impairments. This research mainly develops an intuitive platform that offers a seamless way of communication through real-time speech-to-text transcription with an integrated Android interface and multiuser adaptability. It helps speakers with diverse speech patterns by leveraging the use of robust machine learning models, including Convolutional Neural Networks (CNNs) and reinforcement learning, to ensure accurate speech recognition. The adaptive mechanisms will ensure the continuous improvement of transcription accuracy through the interaction with the user, and cloud integration allows scalability in data management with seamless updates. Through its focus on communication barriers, this research aims to empower people with speech disorders, support autonomy, and facilitate inclusive participation in health, education, and everyday life.