My thought was: sender creates a request for sending a file based on the senders id the request is opened and account awaits sender to send a file once a file is successfully sent it is tagged with the senders id
This all happens under one master account.
The master account holder runs a server application which checks periodically if new files have been uploaded.
If so, it will download the files and do with them whatever it pleases. This is quite simple as the link is an http url.
As a matter a fact, we've recently implemented such an "HTTP-based API" as a web service. Whether it's used by servers, client computers or both, it's the perfect automation solution for computers required to transfer files to one another. For those interested in implementing this new web service, please contact us.