Applied AI Engineer - Codex Core Agent

About The Team

The Codex Core Agent team is responsible for developing the foundational components of Codex, focusing on enhancing the agent’s capabilities, accelerating research efforts, and implementing those advancements into production to benefit end users. Their work spans systems that maintain the agent’s operational efficiency, including token handling, latency, cost management, and capacity, along with the core execution processes and interfaces transforming models into effective actions. They also provide infrastructure supporting other teams building upon Codex and establish feedback mechanisms that refine models and agent behavior based on real-world usage.

Role Overview

This role seeks applied AI engineers focused on advancing Codex agents from impressive demonstrations to reliable tools used in practical software engineering tasks. The main objective is to close the gap between research innovations and real-world usability, improving agent performance and functional impact.

Key Responsibilities

• Design and refine agent behavior for real-world coding challenges and complex workflows.
• Collaborate with research teams to develop evaluations measuring agent performance, identifying regressions, failures, and edge cases.
• Enhance performance via improved prompting, strategic tool usage, context management, and model-oriented experimentation.
• Investigate production failures systematically to boost robustness and reliability.
• Develop feedback loops and data mechanisms to integrate real-task data into evaluation and research processes.
• Partner with product teams to shape user-facing agent experiences and related system interfaces.
• Establish criteria defining successful agent completion of complex end-to-end tasks.

Ideal Candidate Profile

• Experience delivering machine learning or large language model (LLM) powered products.
• Proficiency in Python and familiarity with modern machine learning tools.
• Background in model evaluation, fine-tuning, or prompt engineering.
• System-oriented thinking with a focus on user outcomes beyond mere model metrics.
• Enjoys diagnosing complex real-world issues and transforming them into improvements.
• Passionate about bridging research advances and practical user-facing AI systems.

Additional Qualifications (Preferred)

• Experience with agent frameworks or LLM-based tool usage systems.
• Research background in code generation models or software developer tooling.
• Exposure to handling large, unstructured datasets or logs in production environments.

About OpenAI

OpenAI is a research and deployment organization dedicated to ensuring that artificial general intelligence benefits all of humanity. The company pioneers capabilities of AI systems while prioritizing safe and human-centered deployment. OpenAI values diverse perspectives to address the broad spectrum of human experience.

Committed to equal opportunity employment, OpenAI prohibits discrimination based on race, religion, color, national origin, sex, sexual orientation, age, veteran status, disability, genetic information, or other legally protected statuses. Applicable background checks are conducted consistent with relevant laws and ordinances, ensuring fair treatment of candidates with criminal records according to jurisdictional requirements.

Reasonable accommodations are provided for applicants with disabilities upon request. OpenAI maintains a global applicant privacy policy supporting this role and our initiatives.

Compensation

The salary range offered for this position is between $230,000 and $385,000 annually.