Role Overview
The Identity Infrastructure Engineering team builds identity and access systems that OpenAI’s researchers, engineers, and internal teams rely on to protect critical systems and data. We’re looking for a hands-on software engineer to design and build high-scale, high-throughput secure access proxies and the distributed systems behind them. These are tier-zero services: researchers, engineers, and agents depend on them to reach critical infrastructure safely and reliably. You’ll work at the intersection of networking, distributed systems, and infrastructure security, building serving paths and control-plane integrations that connect identities to resources. Teleport is a core part of the current platform; the role also includes credential brokers, policy enforcement, and developer-facing integrations. You’ll own services from design through production, improving throughput and tail latency, making authorization decisions auditable, and designing for safe multi-region rollout and recovery.
Responsibilities
- Design and implement secure access proxies, gateways, and credential brokers providing narrowly scoped, auditable access while reducing standing privileges.
- Integrate identity and access policy across clients, proxies, and target systems; ensure access expires or is revoked when required and handle stale policy or active sessions.
- Build observability for the end-to-end access path: latency, errors, saturation, authorization decisions, and identity attribution without leaking credentials.
- Engineer safe multi-region rollouts, canaries, rollback, failover, and recovery; validate with test environments, load tests, and failure testing.
- Keep core access services reliable, performant, and maintainable; own production support, maintenance, upgrades, and incident recovery.
- Make secure access easy to adopt through reusable APIs, client tooling, diagnostics, and self-service workflows; lead migrations with tested rollback and recovery paths.
Requirements
- Built and operated production backend, proxy, networking, or infrastructure services with demanding scale or availability requirements.
- Practical networking depth: HTTP, TCP, TLS, DNS, routing, and load balancing across real request paths.
- Experience handling performance bottlenecks, partial failures, multi-region deployment, or difficult production changes in a system you personally built.
- Understand authentication, authorization, least privilege, trust boundaries, credential handling, and consequences of access-control failure.
- Strong software engineering fundamentals with Go, Rust, Python, or similar languages.
- Use tests, profiling, telemetry, and reproducible experiments to debug across system boundaries and verify correctness, performance, and recovery.
- Make practical scope decisions, communicate tradeoffs clearly, and follow through from code to production. Teleport or comparable access-platform experience is helpful.