The question of how life could have arisen from non-living molecular systems is not only fundamentally and philosophically interesting, but it requires us to closely examine the common features of life as we know it, and to imagine the possibilities for life as we do not know it. While much progress has been made elucidating possible chemical systems as candidates for early pre-biological reactions, little is known about the true catalytic and informational capabilities of small, prebiotically relevant molecules in compartmentalized, micron-scale systems. We address this by using droplet microfluidics to generate encapsulated molecular systems as protocell analogues. We use automated tools to generate, manipulate and analyze these systems in real time, enriching dynamic populations for rare entities and selecting entities of interest for reproduction, recombination and chemical analysis. Using these techniques, we aim to select compartmentalized systems that will provide insight on the origin of complexity from simple biochemical constituents.