Prompt Engineer, Agent Prompts & Evals

About Anthropic

Anthropic's mission is to create reliable, interpretable, and steerable AI systems. We want AI to be safe and beneficial for our users and for society as a whole.

About the role

As a Prompt Engineer on the Agent Prompts and Evals team, you'll bridge model capabilities with real product experience. You'll work across product teams to build consistent, safe, and beneficial user experiences across Claude's product surfaces. Responsibilities span system prompts, evaluations, model launches, and prompt-testing infrastructure.

Responsibilities

• Design and optimize system prompts for consumer and API products
• Build comprehensive evaluation suites for model quality assurance
• Partner with product, research, and safety teams on feature launches
• Support critical model releases and catch regressions
• Develop frameworks and tools for prompt testing
• Mentor engineers on prompt engineering best practices
• Adapt quickly in fast-paced, rapidly-evolving environments

Required Qualifications

• 5+ years software engineering (Python or equivalent)
• Demonstrated LLM and prompt engineering experience
• Strong evaluation methodologies understanding
• Excellent written and verbal communication
• Multi-project management capability
• Version control, CI/CD, and modern development practices

Preferred Qualifications

• Production experience with Claude or frontier AI models
• Machine learning, NLP, or related background
• A/B testing and experimentation framework familiarity
• AI safety and alignment knowledge
• Track record improving AI performance systematically

Compensation

Annual Salary: $320,000 - $405,000 USD

Logistics

• Bachelor's degree or equivalent experience
• Hybrid: minimum 25% in-office
• Visa sponsorship available

Benefits

Competitive compensation with optional equity donation matching, generous vacation and parental leave, flexible working hours.