Hi! My name is Pranav (PRAH-nuv). I am a Research Scientist at Salesforce AI Research and am part of the Interactive AI Team where we’re building trustworthy, interactive AI agents—the kind that can reason with long-term memory, transform unstructured data into structured insights, and enable better human-AI collaboration through advanced interfaces and research.

 
I completed my Doctorate in the Informatics program within the College of Information Sciences and Technology at Pennsylvania State University. I was part of the Human Language Technologies Lab, led by Dr. Shomir Wilson. My research examined sociotechnical language models through the lenses of trust and ethics, with a particular focus on identifying and addressing sociodemographic biases in NLP models and understanding their societal impacts. My interdisciplinary work spans the domains of Human-Computer Interaction (HCI), Social Informatics, and Privacy in NLP. My PhD thesis (which was awarded the best AI Dissertation at PennState) can be found here. 

I have authored multiple papers published in leading NLP and AI Ethics conferences such as ACL, EMNLP and FAccT. My research has garnered recognition at major venues and has been featured in prominent media outlets, including Fast Company, AIHub, The Hill, and VentureBeat. You can find more on my research's media coverage here.

During my internship at Salesforce AI Research, under the mentorship of Jason Wu and Philippe Laban, I worked on improving Answer Engines and Retrieval-Augmented Generation (RAG) systems, focusing on evaluation methods and designing safer AI systems. I was also part of the Primed to (re)act project, led by Dr. Christopher Graziul, a collaboration with the University of Chicago, investigating language and behavior in broadcast police communications, particularly in relation to minority populations. I have also actively collaborated with numerous researchers from both academia and industry to advance the field of Trustworthy Human Language Technology. Among my notable collaborations is my work with Dr. Aylin Caliskan at the University of Washington, where they explored the cultural implications of harm in NLP models and with Dr. Koustava Goswami from Adobe Research to understand the ethical frameworks of hallucination in language models. Some of my other collaborations also include researchers from leading institutions and organizations such as Hugging Face, Meta, Georgia Institute of Technology, and IBM Research, among others.

I have held several prominent leadership roles in the research community, including serving as an Area Chair for ACL 2025, EMNLP 2025 and as a Program Committee member for AIES 2025, FAccT 2025, FAccT 2024, and TrustNLP 2023. Beyond my organizational contributions, I have also been an active mentor, guiding graduate students through impactful research initiatives. Notably, I served as a Project Mentor for the NSF Careers Program, leading a project titled "Bias and Fairness in Machine Learning: Mitigating Bias in Unstructured Data." In this role, I helped students develop and implement models to identify and mitigate bias in language systems.

Before joining Penn State, I worked at Honeywell Technology Services, contributing to automation, process bots, and application virtualization. I earned his B.Tech in Computer Science and Engineering from Amrita School of Engineering, where I undertook projects in full-stack development and conversational automation systems.

Check out my Scholar profile here!

Follow me on Twitter (now X) for more such updates about my work and research! 😄