This has been left answered for a while so I thought I would pitch in with my guesswork!
Obviously it is not possible to reference sizes of customer environment but it is safe to say that the AWS (i think that is now the basis of BPC) cloud infrastructure provided is in itself extremally scalable. You can probably google search how big the biggest AWS virtual desktop environments are and I suspect they are truly enormous.
The next question then is how scalable is the Blue Prism product itself. That question can sometimes be how long is a piece of string, it will very much depend upon how well you maintain your BP database (how long you keep BP data) and if you adhere to best practices in relation to logging data. Version 7 of the product was a massive improvement on previous versions in terms of scalability - I now suspect if you adhere to best practices in delivery and data archiving you should get up to something approaching 2000 robots before you need to start thinking about splitting environments. In reality most organizations split environments well before that scale is needed for internal reasons related to department/domain/country data segregation.
