About the role
About the Team
The Intelligence & Investigations Engineering team builds systems that detect, analyze, and disrupt abuse across OpenAI’s products. We partner closely with the Child Safety team and cross-functional groups to protect users while advancing OpenAI’s goal of developing AI that benefits everyone.
About the Role
As a Fullstack Engineer focused on child safety, you’ll build data-intensive, AI-powered applications and infrastructure that enable operators and investigators to work effectively and responsibly. You’ll adapt quickly in ambiguous, fast-moving environments to deliver well-crafted, reliable tooling for high-severity safety work.
*Candidates should understand this role involves exposure to sensitive and egregious content.
In this role, you will
- 7+ years of experience as a software engineer
- Prototype, build, and maintain intelligence systems that detect, triage, and enable efficient human review of possible high severity harm
- Work hand in hand with operators and investigators, designing and delivering systems that enable them to do their work faster, more accurately, and more safely.
- Develop across the stack: UIs, services, pipelines, and anything else required to solve the problems we face.
- Interact with partners across Product Policy, Platform Integrity, Safety Systems, and Research
- Contribute to the team’s technical strategy, especially for child safety related tools and systems
- Report on impact in a data-driven fashion
You might thrive in this role if you
- Have a strong software engineering foundation and enjoy owning systems end-to-end—from infrastructure and data ingestion to frontend tooling
- Are energized by working at the frontier of AI capabilities, integrating new models and APIs into practical systems
- Have experience building and operating large-scale data pipelines or search/retrieval systems
- Are proficient in Python and/or TypeScript, and familiar with tools like Spark, Kafka, Flink, data warehouses, and SQL
- Take a product-minded approach, design with user workflows in mind, and iterate quickly based on feedback
- Favor pragmatic solutions and can ship in hacky, low-support environments when needed
- Are comfortable navigating ambiguity and open-ended problem spaces with evolving goals
- Have prior experience working on engineering for high-severity harms