This project aims to develop an automated Speech Language Therapy system for children with speech impairments, addressing the need for consistent, engaging, and personalized sessions. Current methods struggle with therapist shortages, lack of child engagement, and data privacy concerns. Our solution involves creating a customizable tongue articulation simulator that generates 3D models based on audio input, using ultrasound imaging data as a foundation. This system reduces the need for constant therapist supervision, automatically generates clinical reports, and ensures data privacy by focusing on audio rather than video data, making speech therapy more accessible and effective for children.