About me

I am a manager and a senior research scientist in the conversational AI group at IBM Research, New Delhi. My primary interests are in the intersection between Large Language Models (LLMs) and conversational AI. At IBM, I work on systematically injecting knowledge unseen during training into LLMs to enhance it’s performance on domain-specific RAG. In addition to that, I also work on infusing function calling capabilitites into IBM Granite models. I have contributed to IBM Watson Assistant’s Language Understanding codebase – specfically the dictionary based entity extractors, misspelt entity extractor and actions runtime.

Previously, I completed my Ph.D at IIT Delhi working with Prof.Mausam. During my Ph.D, I worked on generalizing end-to-end task oriented dialog systems to unseen knowledge.