[Hiring] AI Task Creator @Athyna

Role Description

We're looking for an AI Task Creator for a US-based AI research company building custom reinforcement learning environments and coding agent benchmarks. If you are systematic and autonomous with Python or Go and Docker, keep reading!

• 100% Remote
• Salary in USD
• LATAM (Brazil, Argentina, or Colombia) - Reporting to US, EST
• Contract, hourly (20+ hours/week)
• Advanced English

This is a contract role that can be part-time or full-time depending on your availability.

Qualifications

• Strong, hands-on proficiency in Python and Go
• Solid understanding of the software development lifecycle in a verifiable, binary-outcome domain
• Comfortable working with the concept of agent trajectories
• Familiarity with common task shapes such as bug fixes and basic service failures
• Experience using AI coding tools as part of your daily workflow
• Familiarity with Docker and containerized test environments
• Ability to debug real codebases without a test suite or a clear error message

Requirements

• Source and pin large, complex open-source repositories that host a real planning challenge for the agent
• Write task prompts that state the problem without revealing the plan or solution
• Design roughly ten binary yes/no rubrics per task
• Generate and review a gold trajectory that passes every rubric
• Build deterministic test suites with pass-to-pass and fail-to-pass coverage
• Own the full task lifecycle from ideation through QC and refinement iterations

Skills that will make you stand out

• Reinforcement learning experience, ideally with long-horizon tasks
• Experience contributing to or maintaining a large real-world open-source repository
• Prior task-creation or data-annotation work for an RL or agent-training pipeline
• Comfort reading unfamiliar languages or codebases beyond Python and Go
• Master's degree in Computer Science from a top-tier university