How would you write efficient queries that can handle large-scale datasets without stalling at Halliburton?
