Software Engineer, Sandbox & Agent Executor

Responsibilities

• Own agent behavior across product surfaces, including quality, safety, variance, failure modes, and tool and harness design.
• Design and evolve prompting, context construction, retrieval, routing, and tool-use strategies for long-horizon workflows.
• Build evaluation systems using statistical signals, distributions, and trends to measure agent behavior and prevent regressions.
• Detect, diagnose, and resolve nondeterministic failures such as hallucinations, partial correctness, instruction drift, and context sensitivity.
• Partner with product and infrastructure teams to provision, isolate, and safely roll out agent workloads in cloud and self-hosted environments.
• Build and deploy agentic products across the stack using TypeScript, Node.js, and React.
• Mentor engineers and establish standards for reliable AI behavior and quality.

Requirements

• 6+ years of professional engineering experience with ownership of complex production systems.
• Production experience with agentic systems, including context engineering, tool use, and evaluation frameworks at real user scale.
• Hands-on experience with sandboxing technology and running agents inside sandboxed environments.
• Experience with Kubernetes or an equivalent platform in practice, including EKS, ECS, or similar, and end-to-end service ownership.
• Strong systems thinking and sound engineering judgment when using AI.
• Experience investigating model behavior and mentoring engineers on agent-focused work.