Software Engineer, Agent Infrastructure
AI Tools
Tech Stack
Agent Workflow
Build systems enabling training and deployment of AI agents. Provide a workspace for AI models to execute code, debug issues, and develop software. Maintain production platform powering Codex, Operator, and ChatGPT tool use.
About the Role
About the Team
The Agent Infrastructure team builds systems enabling training and deployment of AI agents. The team designs scalable training environments where agentic models execute code, debug issues, and develop software. They maintain OpenAI's core platform for agent deployment powering products including Codex, Operator, and ChatGPT tool use.
Key Responsibilities:
- Build custom container orchestration platforms exceeding Kubernetes capabilities
- Develop and maintain FastAPI and gRPC APIs for agentic infrastructure
- Use Terraform for complex infrastructure provisioning
- Collaborate with researchers on novel AI training systems
- Build systems from 0-1 quickly, and then scale them 1,000,000x
Ideal Candidate Profile:
- Deep expertise in large-scale ML infrastructure
- Ability to build systems from 0-1 quickly, and then scale them 1,000,000x
- Performance optimization across distributed systems
- Cloud platforms and Terraform proficiency
- Deep technical expertise in virtualization and containerization technologies (e.g. Kata, Firecracker, gVisor, Sysbox)
Salary: $230,000 - $385,000 annually, plus equity and performance bonuses
Location: San Francisco, CA or New York City, NY. Hybrid model requiring 3 days in-office per week.
Benefits: Medical/dental/vision insurance, 401(k) matching, parental leave (up to 24 weeks), flexible PTO, mental health support, relocation assistance.