The work
You will help evaluate and improve code used to train and test AI models. The work covers systems programming, infrastructure, performance-focused applications, scalability, and reliability.
- Curate code examples and create clear software solutions for model-training datasets.
- Correct and refine code in Python, C/C++, Rust, Go, Java, and JavaScript, including ReactJS.
- Review AI-generated code for correctness, efficiency, scalability, performance, and reliability.
- Build agents and other tools to verify systems-level and infrastructure code, and identify patterns in errors.
- Assess model capabilities across prototyping, architecture and API design, production implementation, monitoring, and maintenance.
- Create ways to automatically verify software engineering solutions and work with technical teams to improve coding results against performance benchmarks.
What it pays and takes
- Pay: Not specified.
- Hours: 20+ hours per week.
- Location: Open to candidates based in the United States.
- Engagement: Flexible, part-time contractor role with an initial one-month term and possible extensions based on performance and fit.
- Requirements: At least three years of software engineering experience and strong skills in systems programming, infrastructure, or backend development.
- You should have experience building and deploying scalable, production-grade software, plus strong knowledge of software architecture, design, development, debugging, and code review.
- Practical strength in Python, C/C++, Rust, Go, Java, or JavaScript is required, along with clear written and spoken communication for explaining evaluation decisions.
How it works
Apply on OpenTrain with your resume, then complete the application on the hiring site.