Arhaan Aggarwal
Course Staff, Berkeley Agentic AI
Graduating from UC Berkeley with six majors spanning computer science, statistics and environmental economics and policy, and teaching agentic AI while he finishes
Arhaan Aggarwal graduates from UC Berkeley in spring 2026 with six majors: computer science, data science, statistics, applied mathematics, cognitive science, and environmental economics and policy. He enrolled in 2021 to study computer science and added the rest across four years, declaring the environmental economics and policy major last, in his senior year, because the work he wants to do is policy work. That combination is the one CalCompute keeps needing and rarely finds, since the compute question is technical and political at the same time and few people hold both ends of it.
He teaches the subject as well as studies it. He is on the teaching staff for CS 294/194-196, Agentic AI, run out of Berkeley’s Center for Responsible, Decentralized Intelligence in fall 2025 by Dawn Song, and he was a reader for CS 188, the department’s introduction to artificial intelligence, the summer before.
His research has gone where the stakes are concrete. Through Berkeley’s Data Discovery program in fall 2025 he worked on Hallucinations Detection in Medical Device Question Answering with Mark Shi, Amber Tien and Sean Yue, mentored by Joey Yip at MedHive, on the problem of a model answering a clinical question confidently and wrongly.
The Daily Californian profiled him in November 2025 under the headline “Here to learn”, which is his own account of why he did it. He told the paper he does not use AI tools outside what a course allows, a position worth noting from someone who helps teach the course on building them.
Arhaan Aggarwal is based in Berkeley.