This is going to be an application for both android and iOS however will do POC in android. It is going to have to do the following:
Users will signup onto the website and will be given a reference. This reference will state their name, address and signup time.
User will enter that reference into application, at the same time the app will send to the dB that the reference has been used, will set a cap that each reference can only be used 10times max and only completed once.
User takes selfie
User takes photo of photo ID
User takes photo of address id (eg letter from bank or utility company)
system will then compare:
selfie matches photo on ID
photo ID names matches name per dB
name on address id matches names per dB
address on address id matches address per dB
I want machine learning to do the comparisons so this is then scalable. Please look at faceplusplus for face comparison as see if it is suitable. I am also happy to use other services that have already done the ML work if it reduces your work load.
The backend stack needs to be built on GCP. In quote please state what services you will be using in GCP to complete this. The database holding the signup info will be cloud spanner
let me know if you have any questions