The work
You will create and improve code datasets used to train and test AI systems. The main focus is Rust, with additional work in Python, JavaScript including ReactJS, C/C++, Java, and Go.
Your work will cover software engineering tasks from prototypes and architecture to API design, implementation, monitoring, and operational maintenance. You will also create tools that check code quality, find repeated error patterns, and validate solutions.
- Curate code examples for AI training and code evaluation.
- Develop, correct, and review software solutions.
- Assess AI-generated code for efficiency, scalability, reliability, and benchmark performance.
- Write clear explanations supporting your evaluation decisions.
- Build agents and automated checks that verify coding solutions.
- Work with technical teams to improve AI coding results.
What it pays and takes
This is a remote contractor engagement.
- Pay: Compensation details are handled through standard project defaults; no rate is provided here.
- Time: 20+ hours per week.
- Duration: One month initially, with possible extensions based on performance and fit.
- Work arrangement: Remote, part-time contractor engagement.
- Location: Open to candidates in Austria, Belgium, Canada, France, Germany, Liechtenstein, Luxembourg, Monaco, the Netherlands, Switzerland, and the United States.
- Language: Fluent English for structured written and spoken evaluation work.
- Experience: At least three years of software engineering experience.
- Technical skills: Strong Rust capability, full-stack application development, production deployment, software architecture, design, debugging, and code review.
- Additional programming knowledge: Familiarity with Python, JavaScript, C/C++, Java, or Go.
How it works
Apply on OpenTrain with your resume, then complete the application on the hiring site.
About AI training work
AI training work uses human-written examples and reviews to help artificial intelligence systems produce better results. Experienced software developers are needed to judge whether generated code is correct, reliable, efficient, and useful in real engineering work.