Rogelio Rios

I’m transitioning from being a small business owner and healthcare worker into AI Quality Assurance, bringing a strong focus on accuracy, disciplined process, and “no shortcuts” execution. I’ve trained my approach through hands-on testing—especially adversarial testing (red-teaming) and manual data auditing—to ensure AI outputs match the rules they’re given.

My standout project is “The Policy Guard HR Assistant (2026),” where I built a system to prevent hallucinations by confining responses to a specific company handbook. Using XML tagging and negative constraints, I tested the model to ensure it refused 100% of questions not found in the documentation—demonstrating a practical path to safer, enterprise-ready AI. I also bring operational rigor from healthcare logs and scheduling and from running a small store end-to-end, including data-driven ordering decisions and reducing operational errors in shelf-stocking and inventory counts by 25%.