Is there any design pattern to solve the problem of long running code that uses a lot of integrations?

Let's say you need to write a REST microservice with a single GET request method.

The only function of this method is to receive and aggregate data from different legacy systems through different types of communication channels. These legacy systems are black boxes for us; we cannot change any of them.

As the number of such systems increases, the execution time of the method also increases. And in some cases, the execution time reaches too long values ​​(seconds and tens of seconds).

Hence the question - is there some kind of design pattern or accepted approach for optimizing such services?

Of course, caching can be a solution to the problem. But what if we only have one highly unique input parameter, such as ID? There is no point in making a cache, since the probability of repeating input parameter is very small. What can be done in this case?