The work
Help create and assess code examples used to train and benchmark AI models. The work focuses on Python, with additional coverage of JavaScript, React, C and C++, Java, Rust, and Go.
Review the duties and qualifications before applying.
- Curate code examples, build solutions, and correct code for AI training.
- Evaluate AI-generated code in scenarios involving software architecture, API design, implementation, monitoring, and maintenance.
- Create Python agents and automated checks to test software solutions and find recurring errors.
- Assess solutions for efficiency, scalability, reliability, and code quality, and explain your judgments clearly.
- Work with research and engineering stakeholders to improve AI coding capabilities.
Hours, location, and qualifications
- Location: Open to candidates based in the United States.
- Time commitment: 20 or more hours per week.
- Engagement: Part-time contractor role, initially for one month, with possible extensions based on performance and fit.
- Required: At least three years of software engineering experience and strong Python skills, including production frameworks, tools, and best practices.
- Required: Experience building full-stack applications and deploying scalable software, plus knowledge of architecture, design, development, debugging, and code review.
- Required: Ability to explain technical judgments clearly in writing and conversation.
- Helpful: Experience building production systems at high scale and a strong computer science foundation.
- Helpful: Practical expertise in JavaScript or React, C or C++, Java, Rust, and Go, and familiarity with the full software development lifecycle.