The purpose of this design project is to build a smartphone-based non-mydriatic device that can capture and diagnose a fundus image for diabetic retinopathy and other ocular diseases. The device will consist of a smartphone and an adapter that will include a 45 degree mirror that will be used along with a solid core fiber optic cable to make the smartphone camera and the light coaxial. The image is captured through a 20D lens that magnifies the retina. An Android smartphone application provides a simple UI that allows for image capturing and diagnosis using a trained deep learning image classification model. This product is designed with off-the-shelf and 3D printed PLA components to allow for easy access to income countries which account for 90% of global visual impairment cases.