Distributed Training AI Jobs

Leads the Evals Infrastructure team building distributed systems that schedule and orchestrate evals for frontier models, owning throughput, cost, caching, and researcher-facing harnesses.

Design and build greenfield security systems spanning identity, secrets management, developer/supply-chain security, infrastructure hardening, and secure frameworks like mTLS and cryptographic tooling.

You'll embed within research domains like compute, evals, and RL environments, driving end-to-end execution of ambiguous initiatives, leading large-scale compute planning, standardizing eval readiness for model launches, and owning operational health of RL…

Leads the team building Claude's search platform—indexing, retrieval, ranking and serving—owning search quality, evaluation methodology and platform strategy while balancing product, research and training traffic demands at scale.

Build and improve the RL training infrastructure that researchers depend on day-to-day. Identify and remove bottlenecks across the RL stack: debugging, profiling, and rearchitecting where needed.

Design and run new evaluations of Claude's capabilities — reasoning, agentic behavior, knowledge, safety properties — and produce visualizations that make the results legible to researchers and decision-makers.

Build full-stack platforms for RL environment creation, vendor tooling, and human data collection at scale, including labeling workflows and quality assurance systems.

We're hiring an. Engineering. Manager to lead a. Network. Security. Engineering team in executing across critical initiatives including, but not limited to:. Own the network security roadmap: threat model from the management plane through.

Investigate cross-layer performance gaps against theoretical rooflines, own the correctness evaluation pipeline validating model outputs across hardware and serving configs, and build dashboards for throughput, latency, cost, and reliability.

Working across the full stack to identify and remove bottlenecks preventing progress toward scientific AGI Develop approaches to address long-horizon task completion and complex reasoning challenges essential for scientific discovery.