Contact me: dasarit@ufl.edu
Research
These are my research projects that I was involved in at the Virtual Experiences Reality Group (VERG) Lab at the University of Florida. My work is closely related to the healthcare industries and attempts to find effective solutions for specific users, whether it be young, sexually active African Americans, or clinicians training to regulate their emotions when talking to suicidal patients.
01
Increasing Knowledge and Intentions to Use HIV PrEP among Emerging African American Adults
PI: Benjamin Lok, PhD
Institution: VERG at University of Florida CISE
Project Involvement Time: Fall 2023 - Fall 2024 (1 Year)
Research Focus: Young African American adults face higher HIV rates, and while PrEP is an effective prevention tool, barriers to its use remain. This intervention uses virtual African American humans to educate at-risk LGBTQ+ and heterosexual adults about PrEP, aiming to boost knowledge and intentions to use it, while exploring how interactions with the agents impact health behaviors.
Project Responsibilities: I created virtual humans using Synthesia AI Video Generator to deliver PrEP content and developed the application in JavaScript and HTML/CSS. In Python, I scraped and filtered online HIV PrEP content based on user queries to formulate ChatGPT responses for personalized interactions. I enhanced user engagement by converting chatbot responses into AI-generated audio.
PI: Igor Galynker, MD, PhD
Institutions: VERG at University of Florida CISE, Mount Sinai Suicide Prevention Research Lab
Project Involvement Time: Summer 2024 - Present (4 Months)
Research Focus: When working with suicidal patients, clinicians may become stressed and and respond negatively, which makes communication less empathetic and harms trust with the patients. This could worsen outcomes for the patients, which is why clinicians need training to effectively manage negative emotions towards suicidal patients to improve suicidal outcomes.
Project Responsibilities: I built a JavaScript/HTML system to train the clinicians using AI-generated virtual patients (male teenagers), ChatGPT with Dialogflow for the patient chatbot (communication from clinician to patient), and text-to-speech AI capabilities. I then recorded the interactions into a database for pilot data analysis, which I'm currently doing right now.
02