Principal Software Engineer, Full Stack — Codex

About the job

Responsibilities

• Set technical and product direction for key areas of Codex, translating product vision into scalable architectures and an actionable engineering strategy.
• Lead complex, multi-quarter initiatives involving client applications, backend services, agent systems, cloud infrastructure, and developer tooling.
• Design AI-powered workflows that are generalizable across languages, codebases, development environments, and organizations, while ensuring they are intuitive, reliable, and safe.
• Make critical architectural decisions in ambiguous situations, balancing rapid product learning with long-term system quality and adaptability.
• Identify cross-team opportunities, align stakeholders, and drive initiatives from exploration through launch and iteration.
• Partner with product, design, research, infrastructure, security, and go-to-market teams to translate model capabilities and developer needs into product experiences.
• Engage with developers, open-source maintainers, startups, and enterprises to understand unmet needs and identify foundational opportunities.
• Establish standards for reliability, observability, performance, experimentation, and evaluation for AI-powered developer workflows.
• Act as a technical multiplier through design reviews, mentoring engineers, improving engineering practices, and guiding the team through complex decisions.
• Contribute directly to implementation, especially for novel, technically demanding, or multi-layered problems.

Requirements

• Led the design and delivery of technically complex, user-facing systems with impact beyond a single feature or team.
• Combine deep software engineering expertise with exceptional product judgment, fluently navigating product interfaces, distributed systems, infrastructure, and developer tooling.
• Proven ability to create clarity in ambiguous problem spaces, set technical direction, and align multiple teams or disciplines.
• Remain hands-on while increasing the effectiveness of other engineers through architecture, mentorship, decision-making, and technical leadership.
• Deep understanding of modern software development workflows and the ability to recognize where AI agents require new interaction and system designs.
• Experience building or operating AI-powered products, understanding challenges like nondeterminism, evaluation, latency, reliability, safety, and evolving model capabilities.
• Make thoughtful trade-offs between speed, product learning, technical quality, and long-term leverage.
• Develop strong perspectives from user evidence while being willing to adapt them based on technological and user behavior evolution.
• Communicate clearly across engineering, product, design, research, and executive audiences, especially with substantial uncertainty or organizational complexity.
• Energized by defining problems as much as solving them, and by creating the conditions for others to deliver excellent work.