Remote Research Engineer: Code Gen & AI Model Evaluation

YO AI Labs; Wellington; Remote; NZD 143,000 - 214,000; Full time

Job summary

Remote contractor role for research engineers focused on code generation, software engineering, and AI model evaluation. You will analyse, review, and evaluate real-world coding tasks across multiple languages to help train and improve next-generation AI systems.

You will apply programming expertise to develop features, debug issues, and optimize codebases, while providing clear feedback to support model training and evaluation.

Qualifications

• Strong expertise in competitive programming, algorithms, and data structures.
• Experience debugging, feature implementation, codebase refactoring, and performance optimization.
• Experience with open-source contributions or collaborative software projects.
• Excellent written and verbal communication skills.
• High attention to detail in code validation and output consistency.

Responsibilities

• Analyze, debug, and resolve issues across codebases using Python 3, Java, Rust, C++, Go, or TypeScript.
• Implement new features and enhancements for code generation workflows.
• Refactor and optimize code for performance, scalability, and maintainability.
• Develop and review coding tasks to evaluate AI-generated code.
• Provide clear feedback and annotations to support model training and evaluation.
• Collaborate with technical contributors and stakeholders on software quality and technical direction.
• Document technical decisions, solutions, and best practices.

Skills

Competitive programming

Algorithms

Data structures

Python 3

Java

Rust

C++

Go

TypeScript

Open-source contributions

Job description

Remote contractor role for research engineers focused on code generation, software engineering, and AI model evaluation. You will analyse, review, and evaluate real-world coding tasks across multiple languages to help train and improve next-generation AI systems.

You will apply programming expertise to develop features, debug issues, and optimize codebases, while providing clear feedback to support model training and evaluation.