Anthropic
Performance Engineer, Inference Systems
$350K–$850K
San Francisco, CA +3 · Full-time · Posted May 20, 2026
- Location
- San Francisco, CA
New York City, NY
Seattle, WA
San Francisco, California, United States - Experience
- No fixed number of years stated; years required correlate with internal job level requirements.
- Education
- Bachelor’s degree or an equivalent combination of education, training, and/or experience; field relevant to the role, demonstrated through coursework, training, or professional experience.
- Visa sponsorship
- Visa sponsorship is available, but not for every role or candidate; if an offer is made, Anthropic will make every reasonable effort to obtain a visa and provides an immigration lawyer.
About the job
Investigates and improves inference-fleet performance and correctness across hardware and serving layers, using measurement and analysis to guide high-impact optimizations.
What you'll do
- Investigate throughput, latency, and reliability across system layers; identify root causes and quantify optimization opportunities.
- Improve correctness evaluations and investigate model-output regressions across hardware, numerics, and serving configurations.
- Build observability and modeling tools, and partner with component teams to prioritize and implement optimizations.
What they ask for
- Hands-on performance engineering experience, including profiling, roofline analysis, latency/throughput optimization, and root-cause investigation in complex production systems.
- Proficiency in Python and ability to read, instrument, and contribute to large production codebases.
- Data analysis skills using SQL, pandas, or similar.
- Ability to communicate quantitative findings clearly in writing; interest in correctness, numerics, evaluation design, and regression detection.
Nice to have
- Experience with ML systems, especially training or inference infrastructure and LLM serving; direct large-scale inference experience is a strong plus.
- Familiarity with accelerator performance, reliability engineering, model evaluation, or distributed-systems observability.
Pay
$350,000 — $850,000 USD (base salary)
Tools
Python, SQL, pandas, GPU, TPU
AI focus
Works on inference systems serving Claude, including performance optimization and correctness evaluation of model outputs across hardware and serving configurations.
Model-written summary · the original posting is authoritative