The development of a corpus for Marathi speech impaired patients is an important step towards improving the accuracy and efficiency of speech recognition systems for population with impaired speech. This project aims to create a comprehensive database of spoken Marathi language samples from individuals with different degrees of speech impairment. The collected audio data will be transcribed and annotated with metadata, such as age, gender, and severity of speech impairment, to facilitate further research on Marathi speech disorders. This corpus can also serve as a valuable resource for developing and evaluating new speech recognition algorithms that are specifically tailored to the needs of Marathi speakers with speech impairments. Ultimately, the creation of this corpus has the potential to improve the quality of life for individuals with speech impairments by enabling better communication tools and technologies.