AI Engineer Vihaan Nama on Privacy, Practice, and Empowered Learning

In this episode, we’re visiting Duke University to meet Vihaan Nama, an AI engineer, researcher, and teaching assistant helping shape how AI is taught and built for the real world. From roles at PS&S and JPMorgan to graduate courses on explainable AI and product management, Vihaan brings a rare combination of technical depth and educator insight.

If you’ve ever wondered how to make AI education more human, or how to turn student learning data into actionable insight, personalized support, or even a study partner, Vihaan offers both clarity and concrete examples.

We talk about everything from his early experiments in sentiment analysis to why open-source models matter for student privacy, how retrieval-augmented generation (RAG) is quietly transforming knowledge work, and what schools can do right now to prepare for custom AI tools of their own.

Key Takeaways

-Your Notes, Your Assistant: Vihaan envisions a future where students can chat with their own lecture notes, using LLMs to review, revise, and apply information in their own language and context.

-From Archive to Advantage: Companies (and schools!) are sitting on decades of underused data. With the right AI systems, that information becomes actionable knowledge.

-Trust Through Transparency: Grounding AI outputs in clear, credible sources is key to building trust, especially in high-stakes environments like education and public services.

-Small Models, Big Wins: As open-source LLMs become lighter and faster, even modestly funded schools can host private AI tools, no cloud dependency required.

-Responsible AI = Responsive Leadership: From sustainability audits to ethical guardrails, Vihaan emphasizes that building AI responsibly starts with knowing what your organization values most.