Shital is Research Engineer at Microsoft Research with interests in deep learning and reinforcement learning. His previous works includes co-authoring machine-learned clustering system for web-scale data that powered parts of Bing and co-founding the team. In 2017, he authored AirSim, a physically and visually realistic cross-platform simulator for AI research. In 2018, he authored TensorWatch, a new approach for debugging training and visualization of vision models. More recently he co-authored Archai, a platform for Neural Architecture Search (NAS) that has produced the tiny Transformer based models powering the text completition feature in Microsoft Office. Currently, he is working on math and code generation models with goal towards achieving perfect reasoning.
Follow him on twitter.