I build end-to-end products — from polished React/TypeScript front-ends to Node/Django back-ends — and work across AI/ML and data science to build clean reliable software.
Junior in Computer Science at Purdue University
Graduating Winter 2027 (Dean's List & Semester Honors)
Interested in: Software Engineering, Machine Learning, Data Science & AI
I build and run live products: GetUrSeat, RRender.ai and GuardCMD.
Experience
Data Science Intern — FedExFedEx
Data Science Intern at FedEx
Jun 2026 – Aug 2026 · Memphis, TN
Developed a Python-based machine learning forecasting model to predict package volumes using FedEx sorting logic and operational data
Built an Agentic AI solution with Gemini and Vertex AI for business-insight retrieval and automated AI actions
Contributed full-stack development across the chat UI, backend, web search, and ML integrations
Shipped on GCP, Databricks, and Azure
IT / QA Intern — SedgwickSedgwick
IT / QA Intern at Sedgwick
May 2025 – Aug 2025 · Memphis, TN
QA intern at a global leader in risk, claims, and loss management
Conducted RPA bot testing and authored detailed test plans
Automated workflows with UiPath and ran enterprise application testing with Playwright
Undergraduate Researcher — John Deere — The Data Mine, PurdueJohn Deere
Undergraduate Researcher at John Deere — The Data Mine, Purdue
Aug 2024 – May 2025 · West Lafayette, IN
Built a Python ML model to predict parts demand using Random Forest, XGBoost, and custom polynomial regression
Evaluated performance with RMSE and identified seasonal demand trends
Worked in an Agile/Kanban environment under John Deere mentors
Software Engineer Intern — Roar LabsRoar Labs
Software Engineer Intern at Roar Labs
Sep 2022 – Apr 2023 · Remote (Mountain View, CA)
Built frontend for joinroar.co using JavaScript, NodeJS, Pug, and GitHub
Worked directly under the CEO at a startup empowering people in tech careers
Gained hands-on experience across building, testing, deployment, and collaboration
Projects
Designed, built and deployed end to end.
Live products
Products I build and run.
Frontend, backend, database and cloud. Drag or scroll through the work.
GetUrSeat
Live
A production full-stack web app built with React, TypeScript, Python, MongoDB, and REST APIs — designed, built, and deployed end-to-end across frontend, backend, database, and cloud (Vercel + Railway).
An AI creative platform that turns natural-language prompts into motion graphics, 3D Blender renders, narrated animations, and editable designs. Built with React, TypeScript, Node.js, Python, Remotion, Blender, and FFmpeg — with OAuth/OTP auth, Stripe billing, PostgreSQL/Prisma, AI orchestration, and Railway deployment.
Abuse protection for apps and AI agents. It scores signups, logins, AI generations, and agent tool calls from 0 to 100 and returns allow / challenge / throttle / review / block. It ships as an SDK, an MCP server, and a coding-agent skill that wires protection in for you.
An AI-powered calendar and scheduling app with iCal integration, a self-hosted small language model for smart planning, and per-user data storage — running on my own Railway infrastructure.
Mobile app with a wardrobe to store clothing pieces online. Build and store outfits manually or use AI suggestions. Includes AI to generate a Pinterest-board style image or dress up an AI avatar to preview how you'd look.
Work in Progress
On GitHub
On GitHub
Video AI pipeline
A set of Python tools I built for turning long-form video into short-form content.
Takes a YouTube URL or upload and auto-generates 20–40s highlight clips with burned-in captions, using Whisper and an embeddings-based viral similarity engine.
A modular feature-extraction pipeline: transcripts, loudness, silence, speech rate, pitch, overtalk, and face presence, all time-aligned for highlight scoring.
Build outfits from your own wardrobe. Snap a garment and it becomes a clean ghost-mannequin render, auto-tagged by AI. Compose fits, try them on your own photo, and generate multi-view style frames. One codebase ships iOS, Android, and web.
A context compiler for AI coding agents. It compiles a codebase into an adaptive k-ary context tree with O(1) hash indexes and hot/warm/cold memory, so agents fetch compact 5–12 node context packets instead of re-scanning the repo. It includes a 3D WebGL viewer.
Tracks the open positions of Polymarket's top ~100 traders and surfaces what the smart money agrees on. Given a bankroll, it generates an AI-built spread of trades validated against live web search, shown on a command-center dashboard. It has zero npm dependencies.
A campus marketplace for Purdue students to buy and sell textbooks, furniture, tickets, and electronics. It has auth, listings, chat, and favorites, and runs on the web, iOS, and Android via Capacitor.
It edits prose so the page reads as one person thinking about one subject for one reader. It checks the facts first, then the shape of the reasoning, then sentence rhythm, and gets to vocabulary last.
