Victor Soares Bursztyn is a Research Scientist at Adobe Research’s AI Experiences Lab.

His research on LLMs has been motivated by various applied settings (e.g., conversational assistants, conversational recommender systems, data analytics) and NLP tasks (e.g., question-answering, explanation generation, code generation, user modeling, planning with tools), being published in venues such as ICLR, ACL, EMNLP, CHI, UIST, and VIS, while originating over 20 patent submissions. Importantly, some of his work is currently deployed at the backend of AEP’s AI Assistant and CJA’s Intelligent Captions feature.

Victor was fortunate to intern twice at Adobe Research (2021 & 2022, with some results available here and here) and work under the supervision of Prof. Larry Birnbaum during his PhD at Northwestern University (2017-2022); before that, he was a Research Scientist at Dell’s Brazil R&D Center.

Outside of work, as a Latin-American, Victor is passionate about improving access of underrepresented communities to opportunities in Computer Science & AI. He also loves to spread the word about the wholesomeness of tropical fruits in the hope of increasing Cashew import — the fruit, not the nut! — in the US 🙂

If you have published work in applied NLP and are interested in a Research Science internship over the next summer, feel encouraged to send him a message with your paper(s) and current research interests.

Publications

How Aligned are Human Chart Takeaways and LLM Predictions? A Case Study on Bar Charts with Varying Layouts

Wang, H., Hoffswell, J., Thane, S., Bursztyn, V., Bearfield, C. (Oct. 13, 2024)

IEEE VIS 2024

Representing Charts as Text for Language Models: An In-Depth Study of Question Answering for Bar Charts

Bursztyn, V., Hoffswell, J., Guo, S., Koh, E. (Oct. 13, 2024)

IEEE VIS 2024 (short paper)