Anthropic is a public benefit corporation headquartered in San Francisco, focused on creating reliable, interpretable, and steerable AI systems. The Discovery Team within AI Research & Engineering works on building AI scientists—systems capable of solving long‑term reasoning challenges for scientific discovery. As a Staff Research Engineer, you will work end‑to‑end, identifying and addressing key blockers on the path to scientific AGI, building distributed training and optimization systems, and creating benchmarks for scientific workflows and computer use.
Staff Research Engineer, Discovery Team at Anthropic
Hybrid - San Francisco, CA
More jobs at AnthropicSalary
USD 350,000 - 850,000
Requirements
Skills
- 8+ years of ML research experience
- familiarity with large scale language model training, evaluation, and inference pipelines
- obsessively iterating on immediate blockers toward long-term goals
- thrives working collaboratively to solve problems
- expertise in performance optimization and distributed computing systems
- strong problem‑solving skills and ability to identify technical bottlenecks in complex systems
- ability to translate research concepts into scalable engineering solutions
- track record of shipping ML systems that tackle challenging multi‑step reasoning problems
- expertise with performance optimization for language model inference and training
- experience with computer use automation and agentic AI systems
- history working on reinforcement learning approaches for complex task completion
- knowledge of containerization technologies (Docker, Kubernetes) and cloud deployment at scale
- demonstrated ability to work across multiple domains (language modeling, systems engineering, scientific computing)
- experience with VM/sandboxing/container deployment and large‑scale data processing
- experience working with large scale data problem solving and infrastructure
- published research or practical experience in scientific AI applications or long‑horizon reasoning
Responsibilities
- 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
- Scaling research ideas from prototype to production
- Create benchmarks and evaluation frameworks to measure model capabilities in scientific workflows and computer use
- Implement distributed training systems and performance optimizations to support large‑scale model development
Technologies
language model trainingevaluation pipelinesinference pipelinesperformance optimizationdistributed systemsVM/sandboxing/container deploymentlarge‑scale data pipelinesdistributed training systemsDockerKubernetescloud deployment
See if your resume is ready for this job
See how our AI can optimize your resume and improve your chances for this role.