Staff Software Engineer, Frontier Security Team

📋 Description

• Architect and build the Agentic Harness that executes complex, multi-step AI workflows across
• Design stable interfaces for tool execution, context construction, state management, memory
• Own agent quality end to end by building evaluation harnesses, representative datasets, automated
• Convert ambiguous reports such as “the agent feels worse” into measurable failure modes
• Analyze production agent trajectories to identify failures in reasoning, retrieval, tool use
• Close the loop between production incidents, root-cause analysis, evaluation coverage, and

🎯 Requirements

• 9+ years of software engineering experience, including technical leadership of complex production
• Direct experience shipping and operating LLM applications, AI agents, or model-backed workflows in
• Strong background in distributed systems, service architecture, high-throughput APIs, concurrency
• Experience building an agent runtime, workflow engine, developer platform, evaluation system, or
• Demonstrated ability to evaluate nondeterministic systems without relying on a single aggregate
• Fluency in Python and strong proficiency in at least one systems or application language such as

🎁 Benefits

• Bonus experience: Building evaluation or observability infrastructure for agentic coding, data
• Bonus experience: Designing human-evaluation programs, scoring rubrics, annotation workflows, or
• Bonus experience: Working with multi-agent orchestration, long-running agents, asynchronous
• Bonus experience: Developing synthetic tasks, simulations, adversarial tests, red-team exercises
• Bonus experience: Building retrieval systems that use vector search, hybrid search, semantic
• Bonus experience: Operating multi-tenant systems that process sensitive enterprise data.