Software Engineer, Agent Platform

Responsibilities: Own and evolve the internal LLM agent framework: abstractions, runtime, developer experience, integrations, reliability Provide framework capabilities and architectural guidance to multiple business lines for LLM agents Collaborate with ML teams on model post-training workflows (SFT, RL, environment-driven improvements) Design tooling for modern agent patterns: tool calls, memory, retrieval, planning loops, subagents, agent graphs, human-in-the-loop workflows Define evaluation suites: task success, tool-call correctness, trajectory quality, robustness, deployment readiness Stay current on agent architecture and evaluation trends and assess adoption feasibility

Key requirements: Strong backend engineering experience building production-quality platforms, APIs, or infrastructure Deep expertise in LLM agent framework design and orchestration patterns (linear, graph-based, multi-agent, plan/executor loops, tool-heavy) Experience designing agent evaluation paradigms (trajectory evaluations, LLM-as-judge, rubric-based grading, adversarial testing, human-in-the-loop) Familiarity with model post-training workflows (SFT, preference tuning, reinforcement learning, environment-based training) Strong judgment on reliability, observability, debugging, and safety for high-stakes LLM apps Ability to work with partner teams to translate needs into reusable platform capabilities collaboration with cross-functional teams problem-solving with ambiguity clear communication of complex concepts LLM agent framework design Agent evaluation platforms and test harnesses Kubernetes