Amir's Research Notes

I’m a scientist and a builder with a slightly unusual origin story: I started my career mapping the oldest light in the universe, then shifted my focus to AI when deep learning was just becoming a real thing for solving real-world problems.

I trained first as a quantum field theorist at IASBS and then as an astrophysicist at IUCAA, then Princeton and Universit of Toronto (CITA) spending years studying the Cosmic Microwave Background and co-developing statistical techniques still used by cosmologists today. The leap into AI wasn’t a departure from that work; it was a translation of it. The same instincts that helped me find structure in billion-light-year datasets turned out to be exactly what industry needed.

I’ve since built and led AI research teams at Thomson Reuters and Scribd, and I now work at the intersection of frontier research and real-world deployment. My focus in the past couple of years has been on agentic AI; systems that don’t just process information but reason, retrieve, and act. I speak and write on multimodal AI, responsible deployment at scale, and Canada’s innovation potential. Having learned all of that, I am now exploring the intersection of AI and physics, trying to see if we can use AI to solve some of the fundamental problems in the physical world. This includes modern methods for computation, control, measurement, and simulation to build the next generation of tools for solving problems for humans.

At heart, I’m still a scientist: curious, rigorous, and convinced that the best technology is built by people who first learned to ask the right questions.