AI Training Specialist / Technical Data Annotator (LLM RLHF & Code Auditing)
This role involved expert-level auditing of Large Language Model (LLM) outputs using Reinforcement Learning from Human Feedback (RLHF). I validated the technical accuracy of AI-generated code and responses, focusing on logic, reasoning, and safety. My work transformed complex software and logic requirements into high-quality datasets for AI alignment. • Audited and validated computer code outputs in JavaScript, TypeScript, React Native, and Node.js. • Designed adversarial prompts for model safety and robustness testing. • Evaluated and ranked model responses based on helpfulness, honesty, and harmlessness. • Ensured strict data integrity in all high-density labeling tasks.