Developed an outlier detection and treatment feature for the TCS-2021/Data-Mining-Project preprocessing pipeline, focusing on enhancing data quality and user interaction. The solution incorporated multiple statistical methods, including IQR, Z-Score, Modified Z-Score, and Percentile, to identify outliers in numerical columns. Treatment options such as removal, capping, and replacement with the median were implemented to address detected anomalies. Leveraging Python and Streamlit, the work included interactive UI components and visual previews, enabling users to review and validate preprocessing decisions. This addition aimed to reduce bias in downstream analytics and improve confidence in data preparation workflows through robust data preprocessing techniques.