Models are hungry - they want lots of clean, high quality data. My team designs libraries & runs massively pooled screens to feed the models, and learns from the models, in an iterative process to build better biomolecules.