Startup Idea : Parsable diagram images

Very often, at the end of meetings, some diagram is captured as a photo. Such image remains uneditable on the mobile device, sometimes it's shared across the team but still, there is no easy way to adjust it or extend with additional elements. We would like to solve the problem of useless diagrams. The idea is to take a photo of the diagram, send/upload it to our app with a chosen, editable format like Draw.io or PlantUML and as a response, the diagram in preferable form will be returned. The solution involves: - shapes recognition- text recognition- parsing the output to the preferred formatIt might be built with the use of machine learning algorithms and later tied to some mobile or online application.