Self-Serve GPU Platform Engineer — Cluster Infra

Kindredventures; United States; On-site; USD 180,000 - 240,000; Full time

Job summary

Perplexity is seeking to own a dedicated platform team to build a self-serve compute platform for training and inference workloads. You will manage and optimize a GPU fleet across multiple providers, designing scheduling, placement, and Kubernetes-based orchestration to keep workloads efficiently utilized and resilient.

You will collaborate with inference engineers and researchers to simplify complex infrastructure, enable scalable experiments, and ensure low-latency production inference on the

Qualifications

• Deep Kubernetes experience with custom operators and multi-cluster federation.
• Experience managing GPU clusters at scale with NVIDIA, CUDA, networking.
• Experience orchestrating compute across multiple clouds (AWS, GCP, CoreWeave).

Responsibilities

• Build a self-serve compute platform for training and inference workloads.
• Operate and manage GPU fleet across providers.
• Develop scheduling and placement logic for capacity.
• Support long-running training and production inference workloads.
• Own Kubernetes-based GPU orchestration across clusters.
• Implement fault tolerance and observability.

Skills

Kubernetes

GPU clustering

Multi-cloud

Distributed systems

Go/Rust/C++

Tools

CRDs

Operators

Prometheus

Grafana

Job description

Perplexity is seeking to own a dedicated platform team to build a self-serve compute platform for training and inference workloads. You will manage and optimize a GPU fleet across multiple providers, designing scheduling, placement, and Kubernetes-based orchestration to keep workloads efficiently utilized and resilient.

You will collaborate with inference engineers and researchers to simplify complex infrastructure, enable scalable experiments, and ensure low-latency production inference on the