Software Engineer, Codex Security

Build end-to-end AI-native security products and platforms for Codex Security at OpenAI. Own full-stack features, scalable analysis systems, and workflows that translate security research into intuitive vulnerability detection, validation, and remediation tools. Strong full-stack engineering skills and security curiosity required; security domain experience is a plus.

About the job

Responsibilities

Build end-to-end product features for Codex Security, from developer-facing interfaces to APIs, backend services, and workflow tooling.
Own ambiguous 0-to-1 projects across product pods, platform security controls, customer deployment workflows, and cyber model infrastructure.
Translate security research and model capabilities into intuitive workflows that help teams find, validate, prioritize, and remediate vulnerabilities with high signal.
Design scalable, observable systems for long-running analyses across large codebases and enterprise environments.
Partner with product, security research, infrastructure, and customer-facing teams to iterate quickly from real-world feedback.
Help shape the technical direction and architecture of Codex Cyber as the team and product surface area grow.

Requirements

Excellent software engineer with strong fundamentals and enough range to work across product, backend, infrastructure, and developer tooling.
Enjoy building high-quality product and platform experiences across the full stack.
High-agency, low-drama, and comfortable owning features from concept to production in a fast-moving environment.
Built developer tools, workflow-heavy platforms, security products, internal infrastructure, or product features for technical users.
Strong product instincts and care about making complex technical systems understandable and useful.
Excited by security and AI.

Nice-to-Haves

Direct experience in product security, application security, vulnerability research, defensive security, identity, SSO, or secure developer workflows.
Like ambiguous 0-to-1 work and can turn messy requirements into crisp, scalable systems.

Key Technical Challenges

Combining model outputs, security analysis, and user workflows into one cohesive product experience.
Designing high-confidence validation and remediation flows that distinguish real issues from noisy or speculative findings.
Scaling product and platform infrastructure for large repositories, complex enterprise environments, and long-running analyses.
Creating developer-friendly interfaces and APIs that make advanced security capabilities understandable and actionable.
Building platform controls and internal systems that help Codex Cyber move quickly while meeting OpenAI's bar for safety, security, and reliability.