Staff+ Software Engineer, ML Sampling Path

Anthropic
San Francisco
Workplace: OnsiteFull timeUSD 320,000 - 485,000 annuallyFunction: Software EngineeringExperience: 8+ yearsEducation: bachelorsSkills: ["Communication","Mentoring","Technical leadership"]

Build and operate the backend safeguards services that sit on Claude’s token generation path across all platforms. Own end-to-end latency and reliability with SLOs, error budgets, and incident response. Ship hot-path changes with canary/gradual rollouts, latency gating, and fast rollbacks while improving per-token performance and cost efficiency. Lead technical design reviews, set sampling-path direction, and mentor engineers.

Loading

Loading job details...

Preparing the role view and application actions.

FursaFursa
Anthropic
Anthropic
5 days ago

Staff+ Software Engineer, ML Sampling Path

✓ Verified Job

Canonical indexed version, validated from employer's careers page.

Source: Company careers pageValidated by: Fursa AI
Last checked: 17 hours agoStatus: Live

Job Summary

Build and operate the backend safeguards services that sit on Claude’s token generation path across all platforms. Own end-to-end latency and reliability with SLOs, error budgets, and incident response. Ship hot-path changes with canary/gradual rollouts, latency gating, and fast rollbacks while improving per-token performance and cost efficiency. Lead technical design reviews, set sampling-path direction, and mentor engineers.
Location: San Francisco
Workplace: Onsite
Employment Type: Full time
Job Function: Software Engineering
Seniority: Mid level

Key Responsibilities

  • •Design, build, and operate backend systems that process every token on the Claude generation path, including the streaming contract with API and inference engines.
  • •Own latency and reliability end to end by defining/maintaining SLOs and error budgets and leading incident response and postmortem follow-through.
  • •Ship hot-path changes rapidly but safely using canaried and gradual rollouts, error budget/latency gating, and fast rollbacks.
  • •Drive per-token performance by chasing tail latency and keeping cost flat as traffic, models, and per-request checks grow.
  • •Set technical direction for the sampling path through design reviews and trade-off decisions, and mentor engineers to raise the operational bar.

Pay and Benefits

Salary: USD 320,000 - 485,000 annually
Perks:Paid LeaveEquity

Key Requirements

  • •Design, build, and operate high QPS systems at global scale with production accountability (incidents, outages, postmortem remediation).
  • •Strong distributed-systems foundation, including replication/consistency tradeoffs, failure modes, and SLO management under load.
  • •Design systems for graceful degradation, including slow dependencies, dropped streams, and partially rolled-out deploys.
  • •Ship broad, mission-critical changes such as database migrations, interface changes, or rewrites.
  • •Minimum bachelor’s degree or equivalent education/training/experience relevant to the role.
Experience:8+ yearsLLMDistributed systemsHigh scale systems
Education:Bachelor's
Skills:CommunicationMentoringTechnical leadership
Tech Stack:Distributed systemsSLO managementIncident responsePostmortemsStreaming contractsLLM inferenceTransformer-based models

Company Brief

Anthropic
Develops large-scale AI systems and safety research to create reliable, steerable, and interpretable AI assistants and models for commercial and research applications.
Industry: AI & Machine Learning
Company Size: Enterprise (1,001+ employees)
Growth: Scaleup
Valuation: Unicorn (USD 1B+)
Funding: Series C
Headquarters: San Francisco, United States
Founded: 2021
WebsiteLinkedIn