Senior Code Reviewer for LLM Data Training (Java)

Job Description

Job Description

About the Company

G2i connects subject-matter experts, students, and professionals with flexible, remote AI training work such as annotation, evaluation, fact-checking, and content review. We partner with leading AI teams, and all contributions are paid weekly upon approval, ensuring consistent, reliable compensation.

About the Role

We’re hiring Code Reviewers with deep Java expertise to review evaluations completed by data annotators assessing AI-generated Java code responses. Your role is to ensure that annotators follow strict quality guidelines related to instruction-following, factual correctness, and code functionality.

Responsibilities

Review and audit annotator evaluations of AI-generated Java code.

Assess if the Java code follows the prompt instructions, is functionally correct, and secure.

Validate code snippets using proof-of-work methodology.

Identify inaccuracies in annotator ratings or explanations.

Provide constructive feedback to maintain high annotation standards.

Work within Project Atlas guidelines for evaluation integrity and consistency.

Required Qualifications

5–7+ years of experience in Java development, QA, or code review.

Strong knowledge of Java syntax, debugging, edge cases, and testing.

Comfortable using code execution environments and testing tools.

Excellent written communication and documentation skills.

Experience working with structured QA or annotation workflows.

English proficiency at B2, C1, C2, or Native level.

Preferred Qualifications

Experience in AI training, LLM evaluation, or model alignment.

Familiarity with annotation platforms.

Exposure to RLHF (Reinforcement Learning from Human Feedback) pipelines.

Compensation

Hourly rates are personalized based on your experience level, educational background, location, and industry expertise. You’ll see your specific rate in your contract offer before signing. Rates for technical roles can vary significantly based on these factors and can be re-evaluated for different projects based on your performance and experience with SuperAnnotate.