Hi,
There is no way to natively interact with stored procedures in 2E. But let's step back before deciding that's what you really want to do.
You note that you have processing that could benefit from set based processing. How do you know it would benefit? What has led you to determine that? I'm not saying that there aren't situations where this is the case, it's just that more often than not, you simply don't gain enough benefits, and waste time and effort for little gain.
Pragmatically, I have found little benefit from a performance perspective to moving from traditional RLA from RPG. The main benefit that I have found in moving to DDL has been external interfaces where, for example, better field names are needed for the users because they are using tooling that presents the fields to them and DDS 6 character field names are meaningless. Sure, there might be benefits to large set based processing (perhaps millions of records), but I have some hefty set sizes, and really don't have an issue with RPG RLA.
It's really difficult to offer much because I really have little idea of your specific application and what languages you are implementing. It might not even be performance that is the perceived issue (see, I really don't know what it is you are trying to fix)...
You can write source code to call stored procedures from 2E functions, but you would have to manage the result sets yourself, and is probably too much work, and little benefit. 2E can also generate SQL database access, rather than RLA, so you might want to look at that, but there are some new performance improvements in SQL that have not been implemented in 2E.
Enough from me...
Crispin.