Staff Software Engineer, Applied Training

CoreWeave•Sunnyvale, CA

6d•$207,000 - $275,000

About The Position

CoreWeave is seeking a Staff Software Engineer to join their Applied Training team. This role focuses on solving the problem of AI labs spending valuable research time on cluster setup and operations instead of model training. The team is building a Kubernetes-native research cluster platform and sandbox client for agentic training and evaluation. The goal is to provide customers with research infrastructure comparable to that found in frontier labs. The engineer will be an early member of a small team, contributing to the roadmap, working closely with customers and internal teams, and potentially owning specific projects like the research cluster platform or sandbox infrastructure. For the research cluster platform, responsibilities include designing and building a complete research cluster experience, including CLI, job configuration schema, Kubernetes operators, and daemons. This involves addressing researcher pain points such as code distribution, checkpoint-triggered evaluation, cross-cluster scheduling, and programmatic job control. For sandbox infrastructure, the role involves owning the Python SDK and collaborating with the backend team to enable large-scale RL training runs with isolated containers for agent rollouts and benchmarks. The engineer will also write documentation for running popular OSS training frameworks on CoreWeave to assist customers and work directly with large AI labs to understand their internal supercomputing stacks and incorporate that knowledge into product development.

Requirements

8–12+ years building distributed systems, ML infrastructure, or developer platforms
Real Kubernetes experience: custom controllers, operators, scheduling, CRDs, workload orchestration at scale — not just deploying things to Kubernetes or cluster administration
You understand what makes researchers productive: code distribution matters, fast iteration cycles matter, workflows that don't require becoming infrastructure experts matter
Familiarity with training: how distributed jobs get scheduled, how ranks initialize, what breaks at scale
You've shipped infrastructure that other people rely on daily — not prototypes, production systems
Strong communicator who can work with customers and translate researcher complaints into system designs

Nice To Haves

Experience building internal ML platforms or research clusters at a company doing large-scale training
Familiarity with agentic AI: RL training with rollouts, agent evaluation, sandbox isolation for running untrusted code
Background with Slurm, Ray, or similar workload orchestration, and opinions on where they fall short
Experience with container runtimes, isolation (gVisor, Kata), or serverless platforms
OSS contributions to Kubernetes SIGs, Ray, PyTorch, or similar

Responsibilities

Contribute to the roadmap for Applied Training — figure out what actually unlocks new workloads and what's just nice to have
Work directly and closely with customers, and other teams inside CoreWeave that are building cloud native primitives: compute, storage, networking, etc.
For the research cluster platform: design and build a complete research cluster experience — CLI, job configuration schema, Kubernetes operators, daemons — solving the problems researchers actually hit: code distribution, checkpoint-triggered evaluation, cross-cluster scheduling, programmatic job control
For sandbox infrastructure: own the Python SDK and work in a tight loop with the backend team, enabling RL training runs to spawn thousands of isolated containers for agent rollouts and agent benchmarks at scale
Write documentation for running popular OSS training frameworks on CoreWeave to unblock customers and help them succeed
Work with infrastructure teams and customers directly — the customers are large AI labs running thousands of GPUs; understand how they structure their internal supercomputing stacks and bring that knowledge back to what we build