Pratyush Kumar is researching on Watson conversational systems. He is intrigued by the potential of automated agents conversing with humans and other agents in natural language. He is involved in research on building and analyzing such agents, using a range of methods from deep learning to user studies.
Earlier at IBM, he worked on data analytics for green energy sources: He designed and deployed stochastic optimization methods for managing renewables energy resources, with demonstrable value at an industrial scale.
Pratyush holds Bachelors and Masters degrees from IIT Bombay, and a PhD from ETH Zurich. His PhD thesis focused on formal design methods for cyber-physical systems. Many of his research contributions are in the areas of real-time scheduling and network calculus.
He is equally passionate about abstract theoretical models and methods as he is about building concrete systems. He loves reading books, and is particularly interested in Indian philosophy.