I evaluate AI-generated content for safety, accuracy, and policy alignment across categories such as self-harm, harmful bias, manipulation, and sensitive content.

My work includes structured risk classification, severity scoring, and clear justification writing to support high-quality AI training datasets.

I specialize in identifying edge cases, under-response risks, and policy gaps while maintaining consistency across high-volume evaluation tasks.

Experience includes red team prompt analysis and healthcare (Epic EHR) workflow evaluation for safe and compliant outputs.