Backend Developer - AI Trainer (Germany)

📋 Description

• • Design and implement multi-file backend coding tasks focused on bug resolution, feature development, refactoring, and test generation for AI training datasets.
• • Write clear, precise natural-language specifications that define coding problems, expected behaviors, and constraints for AI model evaluation.
• • Develop and extend unit and integration test suites using testing frameworks such as pytest, Jest, JUnit, xUnit, or Go testing to validate code correctness and edge case handling.
• • Create reference implementations in Python, JavaScript/TypeScript, and optionally Java, C#, or Go that serve as ground truth for AI model assessments.
• • Review peer-generated coding tasks for correctness, clarity, realism, and alignment between specifications and code outputs.
• • Identify ambiguities, edge cases, and potential failure modes in submitted code to ensure tasks reflect real-world software engineering challenges.
• • Ensure strict consistency between task specifications, reference code, and expected outcomes across all deliverables.
• • Navigate and refactor large, complex, multi-file codebases to simulate production environments accurately.
• • Participate in code reviews and production migration scenarios as part of task design, ensuring tasks mirror actual development workflows.
• • Deliver high-quality, production-realistic coding challenges that enable AI models to be trained and evaluated on authentic software engineering tasks.
• • Generate structured, actionable feedback on peer submissions regarding code quality, test coverage, readability, and adherence to specifications.
• • Maintain rigorous standards for test coverage and validation artifacts to support accurate AI performance measurement.
• • Operate autonomously on a project-based schedule, managing own time and workflow without direct supervision.
• • Work remotely using personal tools and infrastructure; no company-provided equipment is provided.
• • Accept project-based compensation paid per deliverable, with rates calculated based on estimated completion time.
• • Maintain eligibility to provide services to other vendors or employers, provided such arrangements comply with their policies.
• • Collaborate asynchronously with internal teams and external contributors across global time zones to deliver consistent, scalable AI training content.
• • Contribute to the ongoing refinement of coding task templates and evaluation rubrics based on observed patterns in peer submissions and AI model behavior.
• • Document design decisions and implementation rationale for each task to support transparency and auditability in AI training pipelines.
• • Prioritize clarity and reproducibility in all deliverables to minimize ambiguity for AI models and human reviewers alike.
• • Adapt rapidly to evolving task requirements and feedback loops from AI evaluation teams.