Senior AI Content Trainer & Evaluator (Contract / Remote) — Independent Contractor / Freelance Platforms
Delivered LLM training signals by grading AI-generated responses against truthfulness, helpfulness, formatting, and safety constraints. Rewrote and humanized outputs to create gold-standard baseline text for subsequent model training and fine-tuning. Performed adversarial red-teaming to uncover hallucinations, biases, and logical fallacies and document findings for improvement. • Prompt engineering across technical writing, creative prose, and logic puzzles to evaluate model behavior. • Evaluation rubric application covering accuracy, tone, and policy/safety compliance. • Generation of high-quality reference text suitable for SFT-style training targets. • Red-team documentation of failure modes to support iterative model refinement.