Hi, I'm Andrew Koller


I'm a developer passionate about creating practical solutions to interesting problems. Starting in physical product design, I developed an eye for aesthetics and functionality, which led me to computer science to understand the technical side of bringing ideas to life. Outside of school, I enjoy working with embedded systems, exploring languages like Python, C, and Rust, and combining design with technology through 3D printing and hardware prototyping.

Welcome to my portfolio!

Education

Bachelor of Science, Computer Science

Maryville University, Expected May 2025

Certificates: Software Development, Data Science

Skills

Proficient

  • Python
  • Docker
  • SQL
  • Jupyter Notebooks
  • Git
  • Fusion360

Learning

  • C (Embedded Systems)
  • Rust
  • Vue.js
  • CI/CD (Jenkins)
  • Three.js

Have Used Before

  • JavaScript
  • PySpark & Databricks
  • AWS
  • Anaconda
  • Node.js

Projects

School Projects

  • CI/CD Pipelines
    Automated builds, tests, and deployments using Jenkins running on an AWS EC2 instance.
  • Express Budgeting App (Group Project)
    Collaborative web application built with Node.js and Express.js to help users create and manage budgets effectively.
  • 3D Print Quality Checker
    Developed a platform leveraging OpenCV machine learning models running in Docker containers to detect and classify common defects in 3D prints.
  • Data Analysis and ML with Databricks & Apache Spark
    Used machine learning models on Databricks and PySpark to analyze and draw insights from different datasets.

Hobby Projects

  • This Website
    Personal site built with Vue3 and Vite, showcasing projects, skills, and contact information, deployed via Vercel.
  • 3D Printing
    I design and print custom lamps, fixtures and other random things.
  • 3D Print Visualizer (WIP)
    On this site there is a page that displays 3D models of lamps I have designed. It uses Three.js and some custom shaders I wrote to mimic the texture of 3D printed objects. I also compressed the models to speed up loading times and delivery.
  • Local LLM Training
    Trained and fine-tuned local language models using llama.cpp.
  • Zen Garden Table
    Automated kinetic sand table designed in Fusion360 and implemented using Arduino, Raspberry Pi, and GRBL.
  • Natural Language to SQL (NL2SQL) Web Application
    Flask and Vue.js app converting natural language into SQL queries using locally-hosted LLMs in Docker containers.

Contact Information

Email:

andrew@avkoller.com
Github|LinkedIn