What this interview will probe
Build and operate the large-scale compute and orchestration platform that powers Perplexity's training and serving workloads, working with Kubernetes, Slurm, PyTorch, and primarily AWS. You'll manage GPU clusters, job scheduling, and the reliability of infrastructure shared across research and product teams. A technical interview would probe distributed training infrastructure, container and cluster orchestration at scale, and how you'd debug a multi-node GPU job that's bottlenecked on networking or scheduling.
ExoForm is not affiliated with Perplexity. This is an independent practice page.