Ratna Kandala

Hello! I am Dr. Ratna. Welcome! Thank you for visiting my profile!

Research Interests

1) LLM Evaluation and Human-Centered AI
I study how large language models behave in complex human-facing settings where outputs may influence trust, judgment, emotion, and decision-making. I am especially interested in evaluation frameworks that move beyond fluency or accuracy to assess reliability, uncertainty, autonomy, sycophancy, and downstream human impact.

2) Computational Affective Science and Behavioral Data Science
I study how people express emotion, well-being, distress, uncertainty, and everyday experience in language. This work connects computational modeling with psychological measurement and human self-report data.

3) Health AI, Mental Health, and Decision Support
I am interested in AI systems that operate in health-relevant and emotionally sensitive contexts, including health misinformation, mental-health-relevant language, emotional support, and advice. A central question in this work is how AI can support understanding and decision-making while preserving uncertainty, human autonomy, and context.

4) Culturally Situated and Multilingual NLP
I study how language technologies perform across cultural, linguistic, and social contexts, especially when meaning depends on local beliefs, idioms, health practices, emotional norms, or low-resource communication settings. This includes work on multilingual health misinformation, culturally grounded AI evaluation, and low-resource conversational systems.

5) Computational Modeling of Human and Biological Systems

Building on my background in Systems Biology and earlier computational neuroscience work, I study human and biological systems as dynamic, adaptive, noisy, and context-dependent. This perspective informs my work on language, behavior, health-relevant data, and LLM evaluation, especially in settings where AI systems must interpret complex human experience and reason under uncertainty. I am interested in connecting language-based, behavioral, and self-report measures with biological, physiological, memory-related, and neural mechanisms.

Educational Background:

I am a computational biologist by background.

I am currently working as a postdoctoral researcher in the EMIC Lab, Department of Psychology, University of Kansas, under Dr. Katie Hoemann. I am working on applying Natural Language Processing (NLP) techniques to multilingual and multimodal datasets to understand how the language people use in daily life relates to emotions, behaviors, relationships, and well-being.

Previously, I was a Ph.D. student in Linguistics at the Department of Liberal Arts at the Indian Institute of Technology Hyderabad (IITH). I worked with Dr. Prakash Mondal in the Language and Cognition Group.

Before IITH, I obtained my masters in English Language Studies from the School of Humanities and Integrated Masters in Systems Biology from the School of Life Sciences, University of Hyderabad (UoH), where I worked with Dr. Joby Joseph, Center for Neural and Cognitive Sciences (CNCS) on understanding the algorithms and computational principles underlying the perception & behavior of the neural network(s) responsible for olfaction in insects.

My research has been presented/accepted at various international conferences across AI/ML, Linguistics, Cognitive Science, and Psychology. Please see my Publications/Talks page for a full list.

Here is a short video explaining my entire Ph.D. work in less than 5 minutes (Presented at the LSA Annual Meet in New York City, Jan 2024):

A little more about me

My official name is Kandala Nirupama Sri Ranga Ratna. Unless we've met during high school, I go by the name Ratna. I was born in Warangal, India, and raised in Hyderabad. I hold a black belt in karate (Shotokan style). I practise Vipassana. My other interests include fitness, music, reading, dancing (Bharatanatyam and, currently, Tango), and traveling.

Here is my CV (Jan 2026).I am always interested to hear about new opportunities or even chat about LLMs, language, and karate.Feel free to contact me (including high school students/undergrads) through email for potential collaborative opportunities.