The "Endo-Shut" project is a groundbreaking endeavor aimed at detecting esophageal bleeding using an extensive dataset. Deep Learning (DL) techniques are at the core of this bleeding detection system, and its implementation is facilitated through the Jetson Nano repository. This project relies on crucial code components, including train.py and imagenet.py for Image -classification tasks, as well as detectnet.py and training.py for Object-Detection from the repository, demonstrating a pioneering approach to enhancing medical diagnostics through DL and specialized hardware resources.