I am a student majoring in Computer Science with a cluster in Political Economy and Development at the University of Rochester. I was awarded the Whipple Science Research and Innovation Scholarship, became a Meliora Scholar (which recognises students interested in the humanistic social sciences), and was the recipient of a Research Innovation Grant (RIG) for my research abilties.
I grew up in Dhaka, the capital city of Bangladesh. My experiences made me passionate about amplifying technology for social good. As such, I am a product manager and developer at RocLab, a tech consulting club at my university. I am a founding web developer and designer for Hestia, a crowdfunding intitiative with members from all over the world. Recently, I published a research paper at the Journal of Emerging Investigators (JEI) using natural language processing (NLP) and AI at large to detect human trafficking. I also published an article about this work for Human Trafficking Search (HTS), a global database for human trafficking resources. I presented my research at ROC Talks, inspired by Three-Minute Talks (TmT).
I am also a Foreign Policy Analyst at the Institute for Youth in Policy, where I collaborate with a diverse team to write non-partisan briefs on the world's most pressing issues. I am currently a Research Assistant at the Human-Computer Interaction Lab where I am working on a political misinformation debunking model using generative AI.
My growing interest in sustainability led me to become a Panda Ambassador for WWF. I am a member of the Working Group for Humanitarian Open Street Map (OSM), which is a organization that maps geospatial data in vulnerable and disaster-prone areas.
Proficient in: Javascript (React.js), CSS/HTML, Java, C, Python
For an AI-powered poltiical misinformation tool, I hosted and stored data for a React.js-based website on Amazon S3 using Python Boto3. I also programmatically created and facilitated surveys on Amazon MTurk to gather political human-annotated data for the misinformation debunking tool. I am currently using LangChain to rank news sources respondents are most likely to trust which will then be used to debunk political misinformation. This is research is conducted by the Human Computer Interaction Lab in Unviversity of Rochester.
Melcourses is a course-scheduling app built by RocLab, a club and tech consultancy group at Univeristy of Rochester. As a Frontend Developer, I am working on the cluster search engine. Clusters are similar to minors in University of Rochester, but with less course requirements. This is made with Django and Next.js. I have also made product documentation for other RocLab websites.
As a foreign policy analyst for the Institute for Youth in Policy, I have been working closely with a small team of analysts to assess global issues through an economic, humanitarian, and political lens. To date, I have worked on eight published briefs concerning the Israel-Hamas conflict, Afghanistan's humanitarian crisis, the Nagorno-Karabakh conflict, diplomacy between Greece and Turkey, Bangladesh's 2024 Elections, Houthi Attacks in the red sea, ISIS-K and Brexit. We've been cited by journals like the International Journal of Humanities Education and Social Sciences (IJHESS) and the Italian Political Science Journal.
Understanding that human trafficking is a crime that is often elusive and complex, I built a tool to assess whether it would be possible to detect trafficking from qualitative data. I published a paper at the Journal of Emerging Investigators (JEI) discussing this tool and was selected to speak at ROCTalks, a research talks event at University of Rochester. For this project, I leveraged statistical tools like Mahalanobis distance, Spearman's rho, p-values, alongside Python's Pandas, GeoPandas, Matplotlib, Seaborn and SciPy libraries. This tool had an accuracy of 94%.