The fusion of fashion and technology is transforming personal style, enabling informed and customized outfit selection. This project introduces a deep learning system using Convolutional Neural Networks (CNNs) within TensorFlow to predict the compatibility of shirts and pants. The system analyzes clothing images to evaluate their harmony in patterns, colors, and styles. A diverse dataset of shirt and pant combinations trains the CNN to extract advanced features like texture, color balance, and style elements. Through iterative training, the model identifies visual patterns that define aesthetically pleasing outfits, enhancing its predictive accuracy. The final application provides users with a percentage-based compatibility score, offering practical style guidance. This tool also has potential for retail, enhancing the shopping experience with personalized recommendations. By merging fashion intuition with data-driven insights, the project promotes confidence in clothing choices and paves the way for a personalized approach to style.