LLM Post Trainer
As an LLM Post Trainer, I was responsible for evaluating and validating AI-generated text outputs for coherence and adherence to ground-truth standards. I identified and documented model reasoning failures, which contributed to refining system reliability. My work included providing actionable feedback to improve AI text generation performance. • Conducted comparison of AI-generated answers versus established answers. • Logged and categorized edge-case failures in AI text outputs. • Suggested targeted data-augmentation strategies to address model weaknesses. • Used internal/proprietary tools to review large volumes of AI assistant responses.