You are going to have to give context.
#1: Why can’t you link to the libraries of the other process and keep them all in a single process? Did you not write it/do you not have access to the source code?
#2: Why can’t you write your data to a file, execute the other process with a command-line telling it where your data is and what to do with it, and have the other process write the result to a file which you can then load? Are there restrictions on how the other process has to get and output data?
#3: If nothing else works you can always use ::ReadProcessMemory() to get the result from the other’s RAM, ::WriteProcessMemory() to send your input data (not necessarily in this order), and ::CreateRemoteThread() to call a function in the other process to make it start processing the data. This is clearly a hack and is very advanced, and obviously only a last resort.
You didn’t give any context what-so-ever so there is really no answer to give. Are both processes your own making? If it is not your process, does it have command-line options? Does it already have to be running or can an instance of it be started when you need (and fed a command line)?
Need more information.
L. Spiro