LLM Agentic Trainer, Business Analyst, AI Safety & Policy Analyst
As an LLM Agentic Trainer, I was responsible for translating business requirements and policy guidelines into structured training data for large language models. My work ensured that model behaviors adhered strictly to safety compliance and user intent standards through the design and execution of robust evaluation frameworks. Feedback loops and gap analysis were continuously orchestrated to prioritize and refine high-impact model training scenarios. • Developed precise training and validation datasets in JSON for LLM supervision • Implemented model evaluation strategies to detect reasoning and safety gaps • Translated policy and business logic into training and fine-tuning tasks • Collaborated across research and engineering to ensure dataset quality met 95%+ accuracy