I'm a Mechanical Engineering major at Lehigh University, transferred from Whitman College, where I built a strong foundation in Computer Science, Physics, and Mathematics. I like understanding systems from the ground up, whether that's writing instrumentation software for a physics lab, architecting a full-stack app, or debugging an embedded sensor late at night.

I enjoy building full-stack projects, experimenting with AI and LLM integration, and designing interfaces that make technically complex tools genuinely easy to use.

• Major: Mechanical Engineering (Lehigh University), with a background in Computer Science, Physics, and Math from Whitman College, GPA 3.963/4.0
• Currently: building computer vision and instrumentation software as an undergraduate research assistant in a physics lab
• Currently learning: React and Next.js production patterns, FastAPI backend architecture, and AI assisted development workflows
• Goal: keep shipping polished, production quality projects and grow as a full stack engineer

DuckPath is an AI powered career guidance platform that turns job titles into personalized learning roadmaps. It started as a Streamlit MVP and grew into a production full stack app with a FastAPI backend and a Next.js/TypeScript frontend, including JWT auth, rate limiting, an LLM based skill extraction pipeline, over 85 percent backend and 70 percent frontend test coverage, and a GitHub Actions CI/CD pipeline.

Uni Tracker is a full stack React 19 and TypeScript app for tracking college applications across 5 to 20 schools, with guided checklists, command palette navigation, and real time theming. It uses Zustand and localStorage for backend free state management, backed by a Playwright end to end test suite and automated CI/CD.

Compass is a Slack based student support agent built with Node.js, Slack Bolt, the Groq API, and Supabase, which helps students ask advising questions, find study partners, and connect with campus resources through AI powered workflows.

ESPI Research Software is a cross platform Python desktop application built with PyQt5, OpenCV, and PyVISA for a physics research lab, replacing an expensive proprietary LabVIEW workflow with a free, in house alternative compatible with hundreds of camera models.

Tutor Match is a single page web app connecting students with tutors, featuring a multi role system for students, tutors, and admins, an intelligent tutor auto assignment algorithm, and end to end lifecycle workflows built with vanilla JavaScript.

CS Commons Door Sensor is an embedded security device on an ESP32-S3 with ultrasonic intrusion detection and a Bluetooth Low Energy notification layer.

Languages: Python, JavaScript, TypeScript, C/C++, Java, HTML/CSS, Haskell

Frontend: React, Next.js, Tailwind CSS, Zustand, Plotly.js, PyQt5

Backend: FastAPI, Flask, Node.js, REST APIs, JWT Auth, SQLAlchemy, Alembic

Databases: PostgreSQL, SQLite, Supabase

Testing and DevOps: pytest, Vitest, Playwright, GitHub Actions CI/CD, Git

AI and LLM Integration: Groq API, OpenRouter, LangGraph, prompt engineering

Hardware and Instrumentation: PyVISA (SCPI), Arduino IDE, Bluetooth Low Energy, OpenCV, embedded systems

Concepts: Full stack development, computer vision, data visualization, object oriented programming, data structures and algorithms, CI/CD

GitHub: github.com/Patrick948-stack LinkedIn: linkedin.com/in/mulikuzap Portfolio: patrick948-stack.github.io/myportfolio Email: mulikuzamachumupatrick@g.com