Frontier ML Task Engineer for RL Environments

DevFixr; Poland; Remote; PLN 165,000 - 331,000; Part time

Job summary

DevFixr is seeking an experienced contract software engineer to craft realistic reinforcement learning task prompts for frontier models. You will modify or fix code in existing repositories, apply patches, and prepare clean, unambiguous instructions and tests that force a single correct outcome.

You will run frontier models against these tasks, evaluate outputs against strict criteria, and help extend RL environments for Python and other languages.

Qualifications

• Real engineering depth from shipping and maintaining production systems.
• Strong judgement; junior developers also valued.
• Proficiency in Python; other languages acceptable (Java, C++, COBOL).
• Ability to write precise, unambiguous specifications in English.
• Excellent written English communication.

Responsibilities

• Write tasks by editing repos, changing or fixing code, applying patches, and cleaning up afterwards.
• Write instructions, test cases and expected outputs with a single correct interpretation.
• Run frontier models against tasks and evaluate outputs against rigorous functional and logical standards.
• Help build RL environments for Python and other languages where needed.

Skills

Python

Java

C++

COBOL

English communication

Job description

DevFixr is seeking an experienced contract software engineer to craft realistic reinforcement learning task prompts for frontier models. You will modify or fix code in existing repositories, apply patches, and prepare clean, unambiguous instructions and tests that force a single correct outcome.

You will run frontier models against these tasks, evaluate outputs against strict criteria, and help extend RL environments for Python and other languages.