Anthropic
Research Engineer, Discovery
$350K–$850K
San Francisco, CA +1 · Full-time · Posted May 14, 2025
- Location
- San Francisco, CA
San Francisco, California, United States - Experience
- 6+ years of highly relevant infrastructure engineering experience; the logistics section says required years correlate with internal job level requirements.
- Education
- Bachelor’s degree or an equivalent combination of education, training, and/or experience; relevant field of study demonstrated through coursework, training, or professional experience.
- Visa sponsorship
- Visa sponsorship is available, but not guaranteed for every role or candidate; the company says it will make every reasonable effort to obtain a visa after an offer.
About the job
Build and optimize infrastructure across the model stack to train, evaluate, and deploy an AI scientist and support long-horizon scientific workflows.
What you'll do
- Design and improve distributed infrastructure, data pipelines, and training and inference systems.
- Develop evaluation frameworks and scalable VM, sandbox, and container architectures for scientific AI tasks.
- Translate research needs into production-ready systems and resolve infrastructure bottlenecks.
What they ask for
- 6+ years of highly relevant infrastructure engineering experience with expertise in large-scale distributed systems.
- Deep knowledge of performance optimization and system architectures for high-throughput ML workloads.
- Experience with containerization technologies such as Docker and Kubernetes, and orchestration at scale.
- Track record building large-scale data pipelines and distributed storage systems.
Nice to have
- Experience with language model training infrastructure and distributed ML frameworks such as PyTorch or JAX.
- Experience with GPU/TPU architectures, cloud platforms, reinforcement learning, or AI research infrastructure.
Pay
$350,000 — $850,000 USD annual base salary
Tools
Docker, Kubernetes, PyTorch, JAX, AWS, GCP, Beam, Spark
AI focus
Infrastructure for an AI scientist: language model training, evaluation, inference, and deployment, including reinforcement learning and safe execution of long-horizon scientific tasks.
Model-written summary · the original posting is authoritative