Software Engineer for AI Training

Role Overview

Build reinforcement learning environments that evaluate an AI model’s ability to solve complex software engineering problems using Model Context Protocol tools. You will create real-world tasks that require agents to discover, interpret, and reason over information from MCP servers, helping improve how AI systems learn, reason, and perform. Prior AI experience is not required.

Key Responsibilities

• Create software engineering tasks involving bug fixes, feature implementation, codebase refactoring, and performance optimization.
• Design reproducible environments, deterministic verification methods, and golden reference solutions.
• Develop evaluations that accurately measure both MCP tool use and software engineering capability.
• Produce high-quality, maintainable solutions for complex engineering problems.

Qualifications

• Proficiency in C++, Python, Java, Go, TypeScript, or Rust.
• Strong understanding of algorithms, data structures, and performance tuning.
• Demonstrated experience debugging complex software issues, developing features, refactoring codebases, and optimizing software for performance and scalability.
• Exceptional written and verbal communication skills and close attention to detail.
• Experience collaborating with cross-functional teams, ideally in remote settings.
• Experience with large-scale distributed codebases, rigorous code reviews, and software best practices is preferred.
• Familiarity with modern AI or machine learning systems is beneficial but not required.

Work Terms

• Remote independent contractor engagement.
• Approximately 15 hours per week.
• Flexible schedule, with choice of working days and hours, including weekends if desired.
• Work is paid per task that meets project specifications. Completion time varies based on experience and workflow.
• Minimum weekly submission requirements apply.

Compensation

$60 to $120 per hour. Compensation is output-based and paid for qualifying completed tasks.

Application Process

• Submit an application and complete screening questions.
• Complete an AI interview of approximately 30 minutes.
• A technical assessment may be requested, followed by hiring manager review.
• Selected candidates should be ready to begin their first tasks within 24 to 48 hours after completing onboarding.