Suparnojit Sarkar

Applied Scientist 1 @ Glance

prof_pic.jpg

Bengaluru, India

I’m an Applied Scientist 1 at Glance, where I work with the image generation team.

My research spans multilingual and multimodality, and fairness and bias. Recent work includes MedAidDialog and IndicMedDialog, multi-turn medical dialogue datasets presented at ACL 2026 workshops, and IndiaCultureVL-Bench, a benchmark for cultural recognition in open vision-language models (in review)

I studied Computer Science at Heritage Institute of Technology, Kolkata.

Outside of research, I build and ship applied ML projects — from retrieval-augmented generation systems to multi-agent applications — a few of which are on my projects page. See my publications for the full list of papers.

I’m always open to research collaborations and interesting projects — feel free to reach out.

news

Jul 06, 2026 Paper IndicMedDialog: A Parallel Multi-Turn Medical Dialogue Dataset for Accessible Healthcare in Indic Languages accepted at the BioNLP Workshop, ACL 2026. :tada:
Jul 06, 2026 Joined Glance as an Applied Scientist 1, working on LLMs, computer vision, and generative AI.
Jul 05, 2026 Paper MedAidDialog: A Multilingual Multi-Turn Medical Dialogue Dataset for Accessible Healthcare accepted at the MeLLM Workshop, ACL 2026. :tada:

selected publications

  1. MedAidDialog: A Multilingual Multi-Turn Medical Dialogue Dataset for Accessible Healthcare
    Shubham Kumar Nigam, Suparnojit Sarkar, and P. Patel
    In Proceedings of the MeLLM Workshop, ACL 2026, 2026
  2. IndicMedDialog: A Parallel Multi-Turn Medical Dialogue Dataset for Accessible Healthcare in Indic Languages
    Shubham Kumar Nigam, Suparnojit Sarkar, and P. Patel
    In Proceedings of the BioNLP Workshop, ACL 2026, 2026