Over the last decade, RL has acquired a large repository of algorithms and techniques for solving the most complex of sequential decision-making problems. However, it is still rare to find RL in real-life applications. In this talk, I will cover some of the simplifications and adaptations that we have made at TCS Research to put a mild flavour of RL into the real-world. I will also talk some of the applications where we have been (un)successful, and the reasons why (not).
Harshad is a senior scientist with TCS Research and a visiting associate professor at IIT Bombay. He has recently been invited to become an associated researcher at RBCDSAI. His research interests are in control algorithms for industrial problems, including supply chain, transportation systems, and energy networks. Harshad holds a BTech from IIT Bombay and MS and PhD degrees from MIT.