About
I have worked in the intersection of technology and government which has taught me to put users first. As a software engineer intern and first employee at Aidy (YC W24) I learned to listen to users and iterate quickly based on their feedback. I also previously worked as a researcher at the Massive Data Institute where I tested the fesability of operating a data collection protocol in the real world. Finally, when working at the Beeck Center for Social Impact and Innovation I gained insight into how technology can often be used in much different ways than what was initially tested and how to design for diverse groups of people.
Work Experience
Skills
Posters and Publications
Here are my publications ranging from CS research papers to tech reviews for government professionals.
SCIF: Privacy-Preserving Statistics Collection with Input Validation and Full Security
Developing a data collection protocol to collect statistics from multiple sources while preserving the privacy of each source.
Risk Assessment of Generating Digital Replicas from Text-to-Image Diffusion Models
Exploring how powerful text-to-image diffusion models are at recreating identities with text prompting
Exploring Rules Communication: Moving Beyond Static Documents to Standardized Code for U.S. Public Benefits Programs
Exploring how to translate legal rules into code for public benefits programs
What the Digital Benefits Network is Reading on Automation
A review of articles in the field of automation and its impact on public benefits.
Check out my latest work
I've worked on a variety of projects, ranging from data science to web dev. Here are a few of my favorites.
PorePal
Identify skin conditions using computer vision then make reccomendations on specific skincare products to use.
TinyGen
Given a Github repo link and a prompt, created an API that would generate a diff that accomplishes the given prompt for the Github repo.
Dynamic-Dialo
Dynamically tracking the dialogue state of a conversation. Tracks all necessary information from a conversation and then asks the user the correct question to get the right piece of information based on the context.
Infrared Image Translation
Translating visible light images to infrared images using CycleGAN to detect pedestrians.
I like building things
I attended 5+ hackathons, where people from around the country would come together and build incredible things in 2-3 days. I always made it a point to build something that has real world impact!
- P
PennApps '24
Philedalphia, PA
Swipe to declutter your databases and reduce your environmental impact by reducing databse usage! - H
Hack The North '24
Waterloo, Ontario
Instantly turn images of forms into live webforms ready to be sent out. - H
Healthcare Hackathon with AI (H2AI)
Washington, DC
Filling out medical forms based on electronic health records using computer vision and LLMs. - T
TreeHacks '24
Stanford University, CA
Creating a better developer experience for hardware engineers working with FPGAs by offloading compute to the cloud.
Get in Touch
Want to chat? Just shoot me an email at jsy37@georgetown.edu and I'll respond whenever I can :)
Forked from dillionverma