• Building AI agents and AI-driven workflows to accelerate product velocity

• Published ConMe (https://arxiv.org/abs/2406.08164), which was recently accepted at NeurIPS 2024 Datasets and Benchmarks Track

• Utilized LLMs and VLMs, including GPT4V, with in-context-learning and prompt engineering to evaluate and improve upon Compositional Reasoning benchmarks for Vision-Language Models (VLMs)

• Taught a recitation section of 20+ students twice every week and hosted weekly Office Hours to assist students, for the course 18.C06 Linear Algebra and Optimization; recorded a midterm review video which totaled 140 times and 2398 minutes viewed

• VCMM (Virtu Client Market Making) desk

• Utilized Large Language Models to automate the sensemaking pipeline of conversations for Cortico (https://cortico.ai/), particularly prompting models from the GPT-3.5 family to perform topic tagging

• Contributed as a UROP research student with the Center for Constructive Communication, with research contributions towards the paper "Fora: A corpus and framework for the study of facilitated dialogue," accepted to ACL 2024 (https://aclanthology.org/2024.acl-long.754.pdf)