Staff+ Software Engineer (Safeguards Review Tooling)

Location: San Francisco, CA (hybrid; expected in office at least 25% of the time)

Salary: $320,000–$485,000 USD annually

Company

Anthropic builds reliable, interpretable, and steerable AI systems with a focus on safety and beneficial outcomes.

What you will do

• Build investigation, review, and enforcement tooling across first-party and third-party platform surfaces (case queues, investigation views, decision/audit logging, account-actioning workflows).
• Develop the platform layer for reusable APIs, data storage, and backend services to stand up new review workflows quickly and safely.
• Scale review through automation, enabling reviewers to use Claude and moving toward Claude-assisted and Claude-driven review workflows with humans in the loop.
• Partner with policy, operations, legal, privacy, and data science stakeholders to translate enforcement/investigation needs into reliable systems that measurably reduce handling time and decision error.
• Implement guardrails for sensitive internal tools: granular permissions, audit trails, data-access controls, and reviewer wellbeing features (e.g., content obfuscation, exposure limits).
• Instrument shipped tools with metrics on queue health, reviewer throughput, and decision quality, evolving tooling alongside privacy primitives and data retention commitments.

Requirements

• Technical background in full-stack or platform engineering, with the ability to engage deeply in architecture and design discussions.
• Experience shipping internal tools or platforms for demanding operational users, with a track record of measurably improving workflows.
• Cross-functional experience with non-engineering partners (e.g., operations, policy, or legal).
• Excellent communication skills, including explaining technical tradeoffs to non-technical stakeholders.
• Strong motivation to make AI systems safer and consider societal impacts.

Nice to have

• 8+ years of industry software engineering experience.
• Experience building trust & safety, integrity, fraud, or abuse-prevention tooling, or similar human-review-at-scale systems.
• Experience designing systems under strict privacy/compliance/data governance constraints (e.g., zero data retention environments).
• Experience integrating LLMs/agentic systems into operational workflows, including human-in-the-loop automation and agentic coding tools (e.g., Claude Code).
• Experience building developer platforms or extensible tooling frameworks used by other teams.

Culture & Benefits

• Hybrid policy: staff expected to be in one of the offices at least 25% of the time (some roles may require more).
• Visa sponsorship available; not guaranteed for every role/candidate.
• Competitive compensation and benefits, optional equity donation matching, generous vacation and parental leave, and flexible working hours.
• Collaborative office environment and frequent research/engineering discussions.

Hiring process

• Application review with attention to candidate fit for safety-focused tooling and cross-functional collaboration.
• Visa and immigration support provided via an immigration lawyer if an offer is made.