Question d’entretien chez Meta

How would you handle processing large datasets in a constrained environment?