Staff+ Software Engineer, Product Sandboxing

About the role

Anthropic is building a secure execution environment that powers Claude's agentic capabilities such as code generation, web browsing, and computer control. The team owns the infrastructure that isolates untrusted code and ensures it runs safely, quickly, and reliably for all Anthropic products.

Key responsibilities

• Design and implement a scalable, secure sandbox platform for executing untrusted workloads.
• Lead multi‑month projects, define architecture, and drive cross‑team alignment.
• Collaborate with research and product teams to productize cutting‑edge AI capabilities.
• Own the end‑to‑end lifecycle of services, from design through deployment and operations.
• Ensure the platform meets performance, reliability, and security requirements at massive scale.

Required profile

• Minimum 8 years of experience as a backend or platform engineer building large‑scale distributed systems.
• Proven ability to make architectural decisions and lead technical initiatives.
• Comfort operating in ambiguous, fast‑moving environments and delivering end‑to‑end solutions.
• Product‑focused mindset with a strong sense of ownership.

Required skills

• Proficiency in Python, Go, or Rust.
• Experience with cloud platforms such as GCP, AWS, or Azure.
• Hands‑on knowledge of container orchestration (Kubernetes) and multi‑cloud networking.
• Solid understanding of service‑oriented architectures, networking, and systems design.