The MNIST Handwritten Digits Classification project uses Python, TensorFlow, Keras, and Streamlit to classify 28x28 images of handwritten digits. A trained CNN model predicts digits, and a Streamlit-based frontend allows users to upload images for real-time predictions. This project highlights deep learning's application in image classification