If you have a fixed time window and you want to estimate the number of events, that's a Poisson distribution.

What if instead, you'd like to estimate the amount of time it will take to see a given number of events? Is that still Poisson or is it something else?