Thursday, November 12, 2026
12:00 PM – 1:00 PM PST
Olexandr Isayev: Chemical Experiment at the Age of AI, Agents and Automation
About this event
A free AI for Science seminar at UC Berkeley where Carnegie Mellon chemist Olexandr Isayev explains how machine learning, reinforcement learning and automated flow synthesis are being combined to run chemical experiments with little human intervention.
Part of the AI for Science Seminar Series run by UC Berkeley's BIDMaP. Olexandr Isayev of Carnegie Mellon University presents a machine-learning-guided platform with automated flow synthesis that has been used to search for 19F MRI contrast agents and tough polyurethane elastomers, with reinforcement learning from human feedback driving the optimization.
The talk also looks ahead to LLM-powered agents that plan experiments and read the literature, pointing toward largely autonomous experimental platforms.
In person at the Gateway building, room 1420. Registration is free and asks for name, affiliation and LinkedIn profile. Noon to 1 PM Pacific.
Organized by
UC Berkeley's Bakar Institute of Digital Materials for the Planet, which runs the AI for Science Seminar Series on AI, automation and materials and chemistry research.