Upper limit calculation with very low signal and background level

[eckerpatrick]

Hi,
I am trying to set an upper limit on a very rare decay. My setup is as follows.

1. There is a very low expected background level taken from MC, with the extreme case that the expected background can even go to 0.
2. The signal efficiency is known.
3. I then want to count the observed events which also is very likely to be 0 (or at least a very small number).

As my previous attempts all failed I wanted to ask if it is possible to calculate upper limits on the number of signal events for such a scenario using pyhf and in case how such a model would look like.