About Me
Hey! I am an Honors Computer Science student at UT Arlington researching AI, Computer Vision, and Full-Stack Systems Engineering. My work spans event-based vision, pose estimation, and AI accessibility, using tools like TensorFlow, PyTorch, React, and AWS.
I've contributed to research under multiple professors, developed deep learning models for real-time vision tasks, and led peer mentoring in Data Structures and Algorithms. Passionate about human-centered AI, I aim to build technologies that advance accessibility, innovation, and impact.
Technologies
Experience
Where I've been making an impact
AI Native Software Engineer Intern
To be added
Software Engineer Intern
Developed full-stack Django and React features, automated S3 workflows, improved UI responsiveness, optimized backend queries, and enhanced live production performance.
Software Engineer Intern
Developed scalable SaaS modules with React, Django, and PostgreSQL, optimized APIs and UI, improved reliability, and streamlined Agile-based development.
Undergraduate Research Assistant
Researched event-based computer vision using CNNs and ResNet models, improving event data representations and enabling real-time structural failure prediction systems.
Featured Projects
Things I've built and shipped
Slackbot
Built a production-grade Slackbot and realtime Next.js dashboard for Forward-Deployed Engineers that classifies Slack messages with OpenAI and groups them into actionable tickets using vector similarity and strict guardrails. Designed for reliability and scale with sequential message processing, cross-channel context retrieval, and live ticket updates via Socket.IO.
PiSense
Modular backend for running, evaluating, and improving conversational agents with retrieval and safety controls. Includes FastAPI routes for chat/eval/admin, an in-memory RAG prototype with optional OpenAI or local Hugging Face fallbacks, and Alembic migrations for persistence.
AEDAT Stream Viewer
Web app for side-by-side visualization of AEDAT4 event streams and RGB frames. Backend (FastAPI) parses AEDAT files, serves synchronized event windows and frames; frontend (vanilla JS + canvas) renders low-latency event visualization with playback and scrub controls.