Hydrogen-oxidising bacteria are organisms that play an important role in a multitude of microbial ecosystems, yet they remain relatively unstudied. They use hydrogen, oxygen, carbon dioxide and ammonia to form their biomass and are known to produce high-quality protein, with a composition close to animal protein. Considering their interesting metabolic characteristics, they could be used to valorise waste streams of industrial or agricultural processes (which often contain elevated concentrations of carbon dioxide and ammonia) under form of high quality protein for feed or even for human consumption. However, for this system to become an economically viable option it needs to be highly efficient and productive. In this work we will isolate new autotrophic hydrogen-oxidising bacteria and their heterotrophic commensals from a natural sample. Consequently we will make synthetic in vitro microbial communities by combining one species of hydrogen-oxidising bacteria with different combinations of heterotrophs using a high-throughput culturing method and we will score their performance by measuring biomass. The effort to get the best performing community will be aided by a predictive modelling approach, giving us more thorough insights in the mechanisms leading to high productivity.