infrastructure
Posted Mar 17Software Engineer, Infrastructure - Core Experimentation
at openai
Seattle, United StatesOn-site
Responsibilities
- - Build high-throughput data ingestion and analytics infrastructure for experimentation, product analytics, feature performance monitoring, and model or product measurement workflows.
- - Lead large technical initiatives and shape the architecture of experimentation and rollout infrastructure used across the company.
Requirements
- ABOUT THE TEAM The Statsig team within OpenAI owns the experimentation, rollout, dynamic configuration, and analytics infrastructure that sits on the launch path for OpenAI products.
- After Statsig joined OpenAI, the team began the next chapter: bringing that platform expertise and infrastructure into OpenAI as the experimentation and rollout foundation for every product we ship.
- Recent SDK and server-side infrastructure work has already produced measurable wins in latency, reliability, memory usage, and compute efficiency across important services.
- ABOUT THE ROLE As an Infrastructure Engineer on the Statsig team, you will build and scale the foundational systems behind OpenAI's experimentation and rollout platform.
- You will work on the distributed control plane, SDK and server evaluation paths, ingestion pipelines, analytics foundations, and operational tooling that make launches safe and measurable at OpenAI scale.
- experience building large-scale distributed systems with strict performance, availability, and correctness requirements.
- - Want to work on infrastructure that directly affects how ChatGPT, Codex, model measurement, ads, business subscriptions, developer products, and future OpenAI surfaces ship.
- About OpenAI OpenAI is an AI research and deployment company dedicated to ensuring that general-purpose artificial intelligence
Additional details
- Our systems help teams ship safely, evaluate product and model changes in production, and make high-confidence decisions from real-world usage.
- Statsig began as an independent company built around experimentation, feature management, and product analytics at scale.
- This is infrastructure with a very direct product consequence.
- Teams working on ChatGPT, Codex, model measurement, consumer experiences including ads, business subscriptions, developer products, and shared platform systems depend on Statsig to evaluate configurations, move traffic safely, ingest experiment data, serve analytics, and roll changes forward or back when production reality demands it.
- We are at a critical point in the platform journey.
- Adoption is accelerating quickly across OpenAI, and the systems that were already important are becoming load-bearing for how the company launches.
- The infrastructure needs to stay fast under sharply increasing evaluation volume, reliable when more services depend on it, observable enough to debug quickly, and efficient enough to support OpenAI-wide scale.
- The next phase is to make those gains systematic: a platform that can absorb rapidly growing product velocity while preserving low latency, data quality, operational safety, and developer trust.
- Based out of OpenAI's Bellevue office, we are a close-knit team that values in-person collaboration, technical depth, operational ownership, and building infrastructure that lets other builders move faster without taking on hidden reliability risk.
- This role is deeply technical and centered on performance, scalability, reliability, and correctness.