Georgia Institute of Technology
Anirudh Sundar’s research spans multiple facets of artificial intelligence, with a particular focus on enhancing machine understanding and generation of human language through advanced computational models. His work addresses challenges in question-answering systems, speech recognition, conversational agents, and multimodal interactions, often leveraging large language models and statistical methods to improve performance and applicability in real-world scenarios. In particular, he is interested in situated dialogue over structured knowledge sources such as tables, having worked on retrieval augmented generation using tabular information, generating tables from text via conditional question answering, and presenting a suite of techniques to address interactive dialogue encompassing interpretation, modification, and generation of tabular content using natural language commands.