Keshav Kolluru

About Keshav Kolluru

As a Research Scientist at KnowDis, I work on solving large-scale NLP problems relevant to the E-Commerce domain. I have contributed to a number of systems, including Semantic Product Search, Query Intent Classification, Romanized Indian Language Identification, Machine Translation, Open-Ended Query Segmentation, Advanced Category Classification, and Spelling Correction, many of which have already been deployed and serving millions of users daily. As a Research Scientist at Molecule AI, I work on bringing the advances in Generative AI to the world of drug discovery. I have been involved in developing novel solutions for important problems in the drug discovery pipeline, including de-novo generation - small-molecule ligands and antibodies, ADMET property prediction, molecular docking, and lead optimization. During my PhD at IIT Delhi, I worked on Natural Language Processing with a particular focus on Open Information Extraction and Knowledge Bases for Indian languages. I was fortunate to be advised by Prof. Mausam (Founding Head, School of AI, IIT Delhi) and Prof. Soumen Chakrabarti (IIT Bombay). My thesis research is published in ACL and EMNLP, top-tier NLP venues. I also had the opportunity to work on interesting NLP problems in the industry during internships at Google Research, IBM Research and Microsoft IDC.

Experience

1. Applied Scientist

   Amazon

   Apr 2025 — Present · Gurugram, IN