Srivarsh Vinnu

I am a results-driven engineering graduate with 2+ years of experience in US-IT recruitment and hands-on exposure in AI and LLM evaluation. I specialize in prompt testing, quality assurance, and NLP framework evaluations that improve model accuracy and user relevance.

At Soul AI I evaluate LLM responses and provide continuous, actionable feedback to improve accuracy and safety. I contributed to Project Shield, a side-by-side benchmark comparing two AI agents on reasoning, fluency, relevance and safety, and helped identify hallucinations, biases and inconsistencies. My work blends technical prompt engineering with model performance analysis and stakeholder-facing reporting.

I am adaptable, detail-oriented, and ready to apply my combined recruitment and AI evaluation experience to cross-functional, human-centered projects.