I am motivated to learn as much as possible about all areas of Computer Science. This led me to obtain a wide breadth of experience, including working on AI systems, data ingestion pipelines, distributed systems, embedded systems, web development, and more.
I built a Python-based voice agent pipeline using Azure STT/TTS, GPT, and Retrieval-Augmented Generation over Azure AI Search to prototype a generative AI system for IT Helpdesk requests. The project explored how voice and agentic AI could reduce manual routing latency and improve task automation across internal support systems.
At Aqualab, I developed Splashv2, a large-scale traceroute ingestion pipeline using Dagster, Docker Compose, ClickHouse, and Grafana. The system reduced setup time from days to under an hour and processed 1.6M+ RIPE Atlas traceroutes in under an hour while accelerating geolocation by 86% through multi-core parallelization.
I worked on a peer-to-peer data processing platform built with libp2p, implementing localization to parse UWB sensor data into (x, y, z) coordinates. This contributed to a system that enables distributed tracking across privacy-preserving devices.
I collaborated in 6-person team to develop full-stack web application using React, Node.js, and Supabase to help the community track and share environmental impact data, boosting public engagement.
I engineered a low-latency LED control system in C to study firefly synchronization. The setup achieved sub-5ms latency via a UDP pipeline between a laptop and a Raspberry Pi, allowing for precise real-time behavioral analysis in swarm experiments.
I worked with a four-person fundraising team to raise over $9,500 in sponsorships for Northwestern’s annual hackathon. The role involved outreach to industry partners and coordinating logistics.
At GHY, I created an AI-powered tool that extracted Jira tickets from Git commits and auto-generated release notes across repositories. I also automated database backup verification for Aurora using Python, Boto3, and GitHub Actions.