Copy and evaluation solution

Hello developers,

I would like to consult following setup. Only viable and programmable solutions, keep it real.

My idea is similar to forex rebates services. People give me their accounts credentials for "see only". I can then copy, reverse copy and filter the trades they do on various criteria on more accounts. Example: 100 accounts connected, I group them by ten, and reverse copy each ten of them on a different account of mine. Then the best ten prove worthy of actual direct copying, so I change these ten to normal copy on one account..

I would also like aggregated data for both me and "customer" to see, they can see their own trading data (DD, ratios, trade hold time etc.), while I can filter and display all of them, make groups and groups statistics, or see my accounts performance. So accounts monitoring must be accomplished somehow as well.

BONUS QUESTION: Can also smart automatic grouping option be programmed? For instance, once 10 accounts are connected, reverse copy them on one of my accounts from the prepared list. Once any of them is in profit after 20 trading days, change reverse to direct copy. Retire accounts with no trades during 20 days. Things like that...

My other questions are: How would you do that? EAs on both sides? DLL or not? Or everything on web? How will the evaluating/monitoring work? Are automating options doable? What technical equipment will I need? What trading software would be best for this?

I would like to see your solutions and feedback. How would you do that? Is that even viable/programmable? What is the best copying (speed, slippage) method for this case? Have you done something similar before (do you have experience with web/copy trading software)?

Please, briefly introduce yourself and your solution/feedback (basic methods/architecture), potencial pricing as well for creating it.

Show proof of similar work, if you have any (not mandatory).

After I approve someone, please elaborate your solution the best you can (I am not an idiot, but am not a programmer).

Actual development will be at least month+ away, many other things must be explored first. There is also an option this will not be developed at all, if it´s not viable or better project presents first.

Looking forward for application and cooperation.