Meta Interview Question

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