To clarify my question - say we use bulk to pull 20k patients, but the content of the request is not everything we need (e.g. coverage information, appointments, etc). We would need to call each of these resources for each patient (2 resources 20k patients = 40k calls).
Just trying to figure out the best way to architect this to not overwhelm the servers and hit transaction limits, especially if every client can be different.