Lead Researcher & AI Trainer
As Lead Researcher & AI Trainer at Nomadic Soul 25, I conducted daily intensive training of large language models (LLMs) to align outputs with trauma-aware, somatically accurate, and authentic language. I engineered prompts and iterative logic chains to enhance model understanding of nervous system activation and behavioral outcomes. My work focused on translating field-tested, body-first therapeutic practices into AI training data for use in healthcare and coaching models. • Developed and refined RLHF reward models for somatic and recovery-related contexts. • Authored prompt templates and feedback loops to calibrate empathy and tone in AI outputs. • Synthesized clinical and 12-step recovery outcomes into tangible learning objectives for models. • Utilized multiple AI platforms, including Claude ai, ChatGPT, Gemini, and internal tooling, for model evaluation and fine-tuning.