Reliable Language Model Systems for Educational Assessment and Personalized Adaptive Learning.
Reliable LM Agents · Task Alignment · Reasoning Evaluation · LLM-as-a-Judge · Neurosymbolic Methods · Small & Modular LMs · Memory · Continual Learning · NLP for Education
I’m a fourth year PhD student at the QUEST Lab, IISc, working with Prof. Deepak Subramani on reliable language model systems for educational assessment and personalized adaptive learning. My research spans reliable LM agents, task alignment, reasoning evaluation and verification, LLM-as-a-judge and automated verifiers, neurosymbolic methods for grounding and priors, small and modular LMs, memory and continual learning, and NLP for education.
Alongside my PhD, I am also a consulting Research Scientist at Talking Yak, where I co-led the development and launch of their conversational AI product for language learning and interview preparation. During my PhD I also interned as a GenAI Research Intern at Fujitsu Research of India, and previously as Lead Research Intern at Talking Yak.
Before joining for PhD, I worked as a Project Scientist at Indian Institute of Science, Bengaluru. Prior to that, I did a two-year Teach for India fellowship. I graduated from College of Engineering, Trivandrum with an M.Tech in Control Systems. I completed by B.Tech in Electrical and Electronics Engineering from SCMS School of Engineering and Technology, Ernakulam.
I am also interested in teaching, learning how people think and learn, traveling, hiking, art, music, movies and food.
I am open to discussions and interested in exchanging research perspectives. If you are interested in collaborating, feel free to reach out to me via email!