Remote AI Evaluation Engineer — Build & Benchmark RL Tasks

Cross Border Talents; Michigan; On-site; USD 83,000 - 165,000; Full time

Benefits offered by this job

Fully remote

Flexible schedule

Job summary

Cross Border Talents is seeking an AI Engineer to design and evaluate AI-centric software tasks in a fully remote setup. We seek strong software engineers to craft realistic reinforcement learning environments and reference solutions for benchmarking AI agents.

Prior AI experience is not required; focus is on deep software engineering skills, scalable design, and reliable verification. Flexible hours and remote work enable productivity across time zones.

Qualifications

• Proficiency in one or more listed programming languages and solid CS fundamentals.
• Experience debugging complex software systems and refactoring.
• Ability to design and implement scalable, high-quality software solutions.

Responsibilities

• Create realistic software engineering tasks for AI model evaluation.
• Develop reinforcement learning environments using MCP tools.
• Design tasks involving debugging, feature implementation, refactoring, and performance optimization.
• Build deterministic verification and testing processes to evaluate solutions.
• Develop high-quality reference solutions for benchmarking.
• Ensure environments are reproducible and representative of real-world workflows.
• Analyze complex codebases to identify meaningful engineering challenges.
• Document tasks, solutions, and evaluation criteria clearly.

Skills

Programming languages: C++, Python, Go

Algorithms & data structures

Debugging complex software issues

Performance optimization

Remote collaboration

Software engineering fundamentals

Documentation and testing

Tools

Model Context Protocol (MCP)

Job description

Cross Border Talents is seeking an AI Engineer to design and evaluate AI-centric software tasks in a fully remote setup. We seek strong software engineers to craft realistic reinforcement learning environments and reference solutions for benchmarking AI agents.

Prior AI experience is not required; focus is on deep software engineering skills, scalable design, and reliable verification. Flexible hours and remote work enable productivity across time zones.