Software Engineer, Tokens and Prompt Structures

About the role

The Encodings Infra team builds and maintains the libraries that turn text and multimodal data into the format Claude consumes. As a Software Engineer you will own the design, performance, and API stability of these libraries, enabling researchers and engineers to focus on higher‑level AI work.

Key responsibilities

• Maintain and improve encoding libraries used across Anthropic.
• Run experiments to find optimal ways to feed structured data into Claude.
• Design abstractions that hide encoding details while supporting power users.
• Adapt libraries for emerging research directions and ship them to production.
• Optimize encoding performance throughout dependent systems.

Required profile

• 5+ years of software engineering experience, including library or SDK maintenance.
• Familiarity with ML terminology and LLM architecture.
• Experience with complex refactors in large codebases.
• Strong communication skills and ability to collaborate with researchers.
• Results‑oriented mindset with flexibility to take on extra tasks.
• Commitment to the societal impact of AI work.

Required skills

• Python
• Rust
• Tokenizers or other text/data encoding systems
• Performance optimization
• Reinforcement learning or model training infrastructure

What we offer

• Annual salary $320,000–$405,000 USD
• Competitive benefits, equity donation matching, generous vacation and parental leave
• Flexible working hours and hybrid office policy
• Collaborative research environment focused on safe, steerable AI