Sr. Software Engineer - Perf and Benchmarking
Core
Build and improve Kubernetes-native benchmarking services and maintain end-to-end MLPerf Training and Inference workflows for CoreWeave's global AI infrastructure.
Role type
Senior IC software engineer (performance & benchmarking)
Builds
Benchmarking services, MLPerf submission pipelines, and performance data pipelines for a planet-scale GPU cloud.
Domain
AI Infrastructure / High-Performance Computing / Cloud Systems
Deliverable
production ML models | product features | infrastructure
Required skills
Python, Go, Kubernetes, distributed systems, high-performance computing, networked systems, observability (Prometheus, Grafana, OpenTelemetry), GPU systems (CUDA, NCCL, RDMA), model-serving stacks (llm-d, vLLM, TensorRT-LLM)
Preferred skills
Time-series databases, LSM-based storage engines, MLPerf submissions, OSS contributions (llm-d, vLLM, PyTorch), GPU interconnect topologies
Responsibilities
Ingest, store, transform, and analyze performance events across global data centers; Lead design reviews and drive architecture for multi-service workloads; Mentor junior engineers and elevate coding/testing standards; Ensure reproducible and well-documented benchmarking processes.
Seniority
Senior, hands-on IC