Schedule GPU Pods and Drain a Node
Company: Together AI
Role: Software Engineer
Category: Coding & Algorithms
Difficulty: medium
Interview Round: Technical Screen
Overview: Determine which GPU nodes can accept a request and whether every pod from a drained node can be reassigned without moving existing workloads. Preserve deterministic output while handling whole-pod capacity, exact-fit cases, infeasible placements, concurrent pod creation, and the scale trade-off between exact and approximate scheduling.