I want to leverage AI and Machine learning and develop a mobile app in iOS. I plan to start with clothes but branch into other verticals as well (tools, appliances, etc.). The goal is to blend the in-store experience with mobile commerce, on a personal level (e.g. integrated retail). The app will come prepared with recommendations from the start, but should also continue to learn based on customer preferences as they shop with the app. I need to figure out how to tie this in with Major retailer's APIs as well.
Is something like this possible using CoreML? Or is CoreML something that is local to the device and I would need something that is cloud based so I can tie in with Major retailer APIs?
I believe IBM Watson can address your problem the best. It allows to establish a pipeline where device will upload new data to the cloud and updated models will be pushed back to the device. See https://developer.apple.com/ibm/
I tried their SDK, it works pretty straightforward. I had a problem using UK node of Watson, so I had to switch to US node - and it worked perfectly.