AI Systems Engineer, Codex Agents
Core
Building the agent harness and execution loop that enables Codex agents to interpret model outputs, use tools, execute code, and complete long-horizon tasks safely in production environments.
Role type
Senior IC AI Systems Engineer (Agent Infrastructure)
Builds
Production agent harness, sandboxing/isolation infrastructure, orchestration systems, and evaluation frameworks for LLM-based agents.
Domain
AI Systems / LLM Infrastructure / Agent Orchestration
Deliverable
production ML models | infrastructure
Required skills
Rust systems programming, distributed systems, sandboxing/virtualization, LLM application development, agent orchestration, model evaluation (evals), inference optimization, debugging complex production failures, observability and profiling.
Preferred skills
Compiler/runtime performance engineering, GPU systems, open-source infrastructure development, low-level platform experience, post-training feedback loop design.
Technologies
Rust, Python, LLMs, GPUs, cloud platforms, APIs, logs/traces
Responsibilities
Design and build the core agent harness and execution loop; develop sandboxing and isolation infrastructure; build evaluation and debugging systems; run ablations to improve solve rate and reliability; improve observability across the agent stack; build shared primitives for faster and safer agent operations.
Seniority
Senior, hands-on IC
