Staff Software Engineer – Safeguards Review Tooling

Menlo Ventures; San Francisco (CA); Hybrid; USD 320,000 - 485,000; Full time

Job summary

Anthropic is seeking an engineer to join the Safeguards Review Tooling team. You will own and scale internal tools used by safety investigators to understand and act on harms across Anthropic’s platforms.

As an early member of this new team, you’ll build the platform underneath these tools, including analytics, privacy-preserving primitives, and a sandbox for rapid iteration. Collaboration with policy, legal, and data science teams is essential.

Qualifications

• Technical background in full‑stack or platform engineering and architecture discussions.
• Experience shipping internal tools or platforms for demanding operational users.
• Experience collaborating with policy, legal, operations or non‑engineering partners.
• Excellent communication, able to explain tradeoffs to non‑technical stakeholders.
• Interest in societal impacts of AI and improving system safety.

Responsibilities

• Build investigation, review, and enforcement tooling for first‑party and third‑party surfaces.
• Develop the platform layer of reusable APIs, data storage, and backend services.
• Scale review through automation, enabling Claude-assisted workflows.
• Collaborate with policy, operations, legal, privacy, and data science teams.
• Implement guardrails: permissions, audit trails, data controls, and wellbeing features.
• Instrument tooling with metrics on queue health and decision quality.

Skills

Full-stack engineering

Internal tooling

Cross-functional collaboration

Communication

AI safety mindset

Education

Bachelor’s degree

Job description

Anthropic is seeking an engineer to join the Safeguards Review Tooling team. You will own and scale internal tools used by safety investigators to understand and act on harms across Anthropic’s platforms.

As an early member of this new team, you’ll build the platform underneath these tools, including analytics, privacy-preserving primitives, and a sandbox for rapid iteration. Collaboration with policy, legal, and data science teams is essential.