Aura | Multi-Agent AI-Powered Decision Support System

TLDR:
As tech lead at Honda Research Institute Labs' industry-sponsored capstone project with Carnegie Mellon University, I built Aura™, an AI-powered decision support system leveraging a novel multi-agent system architecture using JavaScript, HTML, and CSS. Throughout the six-month research process, I collaborated closely with diverse stakeholders, including C-suite executives and research teams, translating complex technical concepts into actionable insights and MVP deliverables. My role required working cross-functionally to inform product creation, features, and functionality while providing technical perspectives that guided strategic decision-making, experience.
___________________________________________

Description:
My work exemplified the intersection of technical implementation and empirical research. I applied advanced prompt engineering and in-context learning techniques such as a RAG pipeline to develop a multi-agent architecture that captured diverse stakeholder perspectives with contextually rich responses, demonstrating expertise in AI system design and user research methodologies.

By facilitating participatory workshops with clients to understand organization-specific decision-making patterns and concerns, I gained valuable experience in qualitative research methods and translating user insights into product guidelines. This research-driven approach to product development, combined with my ability to rapidly prototype and iterate on complex AI systems, aligns perfectly with any mission-driven organization to create beneficial AI systems that enhance human capabilities while studying their societal impacts.