• Worked on Agent Opus, an agentic AI system that transforms unstructured inputs into multimodal video outputs, from initial concept to final product
• Developed agent workflows for script generation with grounding in real-world data, and contributed to voice cloning and speech generation pipelines with controllable pacing and human-like expression
• Contributed to digital clone pipelines for generating realistic video outputs
• Worked on agent infrastructure, context engineering, and evaluation systems to improve quality and reliability
• Built customer-facing UI components for interacting with AI generation workflows
