Are you going ahead with a framework (ION, retrofit etc)?
Proper recourse in times of failure (network, timeout, device going to sleep etc)
Are you going to talk about architecting the API or just the consumption? If you are going to talk about architecture, then must include how to fetch related models using a filter/include parameter
How to avoid overmodular APIs (continuation of the 3rd point): Avoid multiple loading spinners
And all the best for your talk! Hope to see you there :)
Hey @labsvisua, just went through your proposal. Looks pretty good! I hope you are able to cover this in 40 mins though. Otherwise you'll have to convert it into a workshop.