← All roles

Staff+ Software Engineer, Claude Managed Agents

Anthropic· Engineering & Design - Product
Apply on company site
Hybrid (San Francisco, CA), Hybrid (New York, NY)$405k - $485kFull-timePosted 2mo ago

About the role

Backend and distributed systems engineer role on the Agentic Systems team building Claude Managed Agents platform for hosting, running, and scaling production agents on Claude. Own end-to-end system design from API through operations, partnering with product, research, developer experience, and GTM to define and ship managed agent capabilities.

Responsibilities

  • Scale the platform: design and operate durable session and event storage, sandbox orchestration, streaming, scheduling, and multi-tenant isolation for long-lived, stateful sessions across various environments; optimize reliability, latency, and cost.
  • Evolve the harness and prove it with evals: collaborate with research to update harness design with each model generation; build eval infrastructure comparing harness quality to baselines and customer workloads; maintain high harness performance.
  • Help builders get the most out of Claude: ship capabilities such as outcome-driven execution, multi-agent orchestration, memory, observability and tracing for long-running agents; maximize intelligence per dollar while ensuring safety.
  • Design APIs that outlast implementations: shape primitives like agents, environments, sessions, vaults, and event streams; focus on versioning, ergonomics across API/SDK/CLI, sensible defaults, and robust escape hatches.

Qualifications

  • Have a minimum of 8 years of practical experience as a backend, distributed systems, or infrastructure engineer
  • Have built and operated stateful, long-running, or high-throughput systems in production (workflow orchestration, streaming, storage, container or job orchestration) with rigour about durability, consistency, failure modes, and cost
  • Strong product sense and treat API design as a craft; collaborate cross-functionally to define product strategy
  • Comfortable with 0 → 1 work, ambiguity, and operating in both early-stage and mature environments
  • Use Claude or AI tools as core to software development; have opinions on what makes an agent harness good
  • Take full ownership from design to deployment and on-call operations; improve what you ship
  • Care about building systems that other engineers and businesses love to use, and doing so safely
  • Strong candidates may have: experience with agent harness/framework, AI/ML platform, evaluation/benchmarking infrastructure for LLMs, durable execution/workflow engines, sandboxed code execution, container runtimes, and public developer platforms (APIs/SDKs)

Skills

Required:Backend / Distributed SystemsAPI DesignSystem Design for Multi-Tenant Platforms

Benefits

  • Competitive compensation and benefits
  • Optional equity donation matching
  • Generous vacation and parental leave
  • Flexible working hours
  • Office space for collaboration
  • Safe and inclusive culture
Last updated Aug 26, 2026. Interested? Apply on the company site.

About Anthropic

building reliable, interpretable, and steerable AI systems

Anthropic is an AI safety and research company that's working to build reliable, interpretable, and steerable AI systems.

ai-infrastructure-and-developmentliquid-ai-recruiting-targetsetter-30setter-30-2026-q1wing-list-2026wing-list-2026-giga-stage
Company profileWebsite

More open roles at Anthropic