Niket Tandon

About

Niket Tandon is a Principal Research Scientist at Microsoft Research, Bangalore working on customizing AI copilots with private data. He works on methods to inject guidance into language models. Previously he was a Lead Research Scientist at the Allen Institute for AI in Seattle until late 2024, specializing in research related to feedback-guided reasoning in Large Language Models. He worked at the Aristo team that created AI to ace science exams. In 2016, Niket obtained his Ph.D. from the Max Planck Institute for Informatics in Germany, where he was supervised by Professor Gerhard Weikum. His thesis resulted in the largest automatically extracted commonsense knowledge graph at the time, called WebChild, which has since been applied in NLP applications and Visual Question Answering. His research has been recognized at various venues. Niket has held other positions in the past, including at Microsoft Research, Redmond; Yahoo R&D, Bangalore; and IBM Labs. With an intent to help the society, he founded PQRS research, an organization providing research opportunities to undergraduate students from underrepresented institutes. He actively serves the research community by organizing tutorials and workshops at conferences, and as area chair in reviewing and in panel discussions, all with a focus in NLP on guidance injection (such as knowledge or feedback injection).

Research Interests

Publications

Publications: