Staff+ Software Engineer, ML Inference Path

Details

The Safeguards ML Inference Path team designs, builds, and operates the production infrastructure powering Claude's machine-learning-based safety systems. The role works at the intersection of machine learning, large-scale distributed systems, and AI safety, productionizing classifiers and ML defenses across Claude's platforms, including 1P, Bedrock, and Vertex.

Responsibilities

• Design and build scalable ML infrastructure for real-time safety deployments across classifier and model ecosystems.
• Build monitoring and observability tools for classifier performance, data quality, and system health.
• Collaborate with research teams to productionize safety research and translate experimental techniques into robust, scalable systems.
• Optimize inference latency and throughput for real-time safety evaluations while maintaining high reliability.
• Implement automated testing, deployment, and rollback systems for production ML models.
• Partner with Safeguards, Security, and Alignment teams to deliver infrastructure meeting safety and production requirements.
• Develop internal tools and frameworks that accelerate safety research and deployment.

Requirements

• Proficiency in Python and experience with ML frameworks such as PyTorch, TensorFlow, or JAX.
• Understanding of distributed-systems principles and experience building high-throughput, low-latency systems.
• Experience building automated or self-service deployment pipelines and evaluation infrastructure for independent classifier and model rollouts.
• Experience implementing A/B testing frameworks and experimentation infrastructure for ML systems.
• Results-oriented approach with an emphasis on reliability and impact in safety-critical systems.
• Ability to collaborate with researchers and translate cutting-edge research into production systems.
• Interest in AI safety and its societal impacts.
• Minimum education of a bachelor's degree or equivalent combination of education, training, and experience.
• Five or more years of experience building production ML infrastructure is a strong candidate qualification.

Preferred Experience

• Large language models and modern transformer architectures.
• Monitoring and alerting systems for ML model performance and data drift.
• Trust and safety, fraud prevention, content moderation, or risk assessment.
• Privacy-preserving ML techniques and compliance requirements.

Benefits

Anthropic offers competitive compensation and benefits, optional equity donation matching, generous vacation and parental leave, flexible working hours, and office collaboration space. The role has a hybrid policy requiring staff to be in an Anthropic office at least 25% of the time. Anthropic sponsors visas where possible and makes reasonable efforts to assist with visa applications through an immigration lawyer.