At TAVILY AI, I evaluate AI-generated responses for technical correctness, factual accuracy, reasoning quality, and policy compliance. I also review advertiser-AI conversations and validate campaign recommendations across the marketing funnel.
I create structured annotation feedback and reusable prompt libraries to support consistent model evaluation. My work includes documenting annotation standards and collaborating with research, product, and engineering teams on evaluation methods.
At TWENTYAI, I developed and optimized prompts for enterprise LLM applications, then reviewed and annotated model outputs using structured frameworks. At MOSAIC AI, I audited training datasets, investigated annotation errors, and contributed to improvements in evaluation frameworks.
At CONTIVIO AI, I implemented conversational AI solutions across voice, chat, SMS, and messaging platforms. I also designed AI workflows and knowledge retrieval systems, and configured and validated agents before production deployment.
