There are many clean solutions to solving image diagnostics, but hospital administrators are resilient to change because of incurring costs, skills mismatch and re-training. Our goal was to supplement and streamline diagnostic data interpretation from both an individual approach (patient cases by case basis) and a wholistic approach (understanding reoccurrences in the general population). We achieved our goal by computerizing the detection of brain tumors in MRI image data, specifically through the use of image processing and machine learning technology. Our targeted user for BrainVision are radiologists who want to speed up their ability to diagnosis MRI data. We used 130+ manually identified tumor and non-tumor images to train a deep convolutional neural network and performed watershed segmentation for feature extractions. Also, we implemented a database feature in MongoDB so that each time a radiologist uploads an image, it will help refine the system. This powerful technology has been streamlined into a simple web app user interface so radiologists can easily use BrainVision to assist their diagnostic process.