Antimicrobial resistant infections are global health concern and predatory bacteria such as the Bdellovibrio bacteriovorus species show promise as a therapeutic targeting these ‘superbugs’, such as the problematic Klebsiella pneumoniae KPC. Our interdisciplinary approach, builds on previous in vitro experimental work and mathematical modelling, extends this to an in vivo environment. Using a zebrafish infection model, with bacteria largely contained within the hindbrain ventricle, allows us to study the interactions of the predator and prey bacteria with the hosts’ innate immune system. We show mathematically and experimentally that the introduction of predatory bacteria can lessen the K. pneumoniae burden in the hindbrain, such that the immune system can clear the infection. Using mathematical models, we can predict outcomes for so far untested predator-prey ratios and test possible mechanisms for predator-immune system interactions.