I think if you're trying to make something similar, a good place to start would be the document channel and product channel. The document channel could optionally include a button to import a photo and OCR that instead.

But you probably know more about app development than I do. All of the things Seeing AI does are useful. Which of those seem doable to you?

Here's a totally off-the-wall suggestion I've been throwing around: I really want to see an app that uses Bluetooth to connect to other nearby devices and allows us to find people in crowds. My scenario is two blind people who are both trying to locate each other. If both of them could get direction or distance information from each other, that would be really useful.