Based just on the text, the question you're asking is a programming one, so i'll provide a programming answer. This will be heavily from a .NET perspective since no other technology requirements were stated in the question.
Assuming you don't rely on the results of a previously processed one, then you could simply task out each of the assessments. I'm more of a .NET programmer and this wouldn't be hard to accomplish in C# but i'm not sure what language you're writing in.
In addition there are considerations to take into account like how much processing you're doing, the resources available etc. The heavier the processing for each, the more you are attempting to parallel process and the lighter your hardware the slower it's going to be.
The only way to know for sure is picking a language to work with, making sure it can parallel process tasks, then working on setting up your "parse" code. Test with 1.. test with 10.. test with 1000. See what results you are getting with whatever hardware. Then move to testing it on a cloud setup. If it takes 5 seconds on your machine, and 5 hours on the cloud rig, then try beefing up the cloud rig to speed it up.
As you're increasing the resources assigned to handle the processing, you'll get an idea of the cost to run your proposed system. Adjust volume processed, adjust code, etc. You'll have to make decisions about volume, cost of hardware resources, etc to get your desired result.
What it SOUNDS like is that it'll be fine. Doesn't seem overly complex or data intensive.
Note: I'm not making any assessment on intent, viability of the idea, or legality that may or may not surround this issue. The question is about processing, as is the answer.
I made a mistake, I'll correct it, thank you. – Josefa Alves – 2018-12-26T17:59:45.267
1I'm voting to close this question as off-topic because it is essentially a programming question about concurrency. – RedGrittyBrick – 2018-12-27T18:21:36.067