Aidan Schmitigal
Headliners
A master's capstone project in partnership with the Oregon Department of Transportation, building a discrete-event simulation of rural highway disaster response across Clatsop County's 3,000+ mile road network. The goal is to optimize how ODOT allocates limited equipment and personnel across landslides, floods, wildfires, and snowstorms first through benchmarked heuristic policies and a policy-switching architecture, then through a trained reinforcement learning agent. Read more
Selected through a competitive process to collaborate directly with Civil Engineering researchers at Caltech, building a browser-based 3D analysis tool for visualizing high-rise building behavior during scenario earthquakes. The tool animates full structural simulations in real-time with displacement, velocity, acceleration, interstory drift data and synchronized panels, threshold flagging, and a custom binary data format I designed to make 8GB simulation datasets actually usable in a browser. Read more
A native Java implementation of Proximal Policy Optimization running directly inside a Minecraft server with no Python, no external processes, no overhead. Using JavaCPP to access LibTorch (PyTorch's C++ API) from Java, the training loop runs synchronously with the server tick, sharing memory with the game engine directly. A dream I've had since I was 14, finally built. Read more
A deep learning final project exploring Transformer-based multi-variable weather forecasting across 145 locations in the western United States. I designed and trained two models. First, a focused single-location baseline and a spatially-aware multi-location Transformer. Secondly, using 5 years of hourly data from the Open-Meteo API, trained on OSU's HPC cluster with CUDA. The project is an honest look at what Transformers can and can't yet do for weather prediction. Read more
A year-long senior capstone project developing a CNN-based system to replace traditional Digital Image Correlation (DIC) for analyzing material deformation in Scanning Electron Microscope images. I led the ML development and designed a synthetic data pipeline that transformed 800 seed images into over 1.4 million unique training examples, enabling a U-Net model to produce per-pixel dense displacement fields, a fundamental improvement over DIC's region-based approach. Read more
Large Projects
Flashcard Royale is a web-based auto-battler game developed in 24 hours at a hackathon. Players can import flashcard decks from Quizlet to create unique battling card units. Correctly answering flashcard prompts powers up card attacks in battles against opponents, blending educational content with engaging, strategic gameplay. Read more
Developed a PDF Region Exporter tool that allows users to export specific regions of a PDF document as a separate page. The tool uses the React framework and the TLDraw library to interact with the PDF.js library for PDF parsing and rendering. The TLDraw canvas allows drawing, editing, and adding extra images to the PDF. A custom region-select tool makes it easy to select pages within the PDF. It also uses the Vercel platform for hosting and deployment. Read more
This is a generative art project that explores the intersection of code and 3D modeling by procedurally generating plants in Blender. Utilizing Blender Python scripting and a custom-designed Geometry Nodes system, this project automatically creates an endless variety of unique, visually captivating digital plants, rendered daily and served as dynamic, automated wallpapers via a web API, Cloudflare Workers, and iOS Shortcuts. Read more
Personal data visualization project transforming YouTube watch history into an interactive dashboard. Reveals viewing habits, content patterns, and platform usage through insightful charts and stats. Designed to make interpreting large and complex user data in a visual way that is easy to understand and digest. Built with React and Recharts, and built with Vite to be a single file, it provides a fast and efficient user experience. Read more
Progressive Web App Sudoku game built with SvelteKit, providing offline functionality and app-like installability. Designed to be mobile-first, with a clean and intuitive user interface. Includes a multi-level hint system, a customizable difficulty level, and a built-in timer. Uses local storage for game state and user preferences. And you get confetti when you win! Read more
Created a grocery shopping planner website using vanilla JavaScript and HTML to visualize the store layout and generate optimized shopping lists for efficient grocery shopping. Created a Trader Joe's price planner website with integration of an unofficial API to enable searching for items and price tracking. Includes an LLM-powered grocery shopping planner that generates list suggestions based on meal ideas. Read more
Full-stack web application to simplify Minecraft server setup and management, automating server creation, deployment, and DNS configuration. Includes programmatic Docker container orchestration for scalable and reliable server hosting. Integrates Cloudflare API for seamless DNS configuration. Uses SvelteKit for frontend-to-backend communication, running on Node.js. Read more
Full-stack family recipe website with customizable recipe books and user accounts. Built with SvelteKit and PocketBase, it creates a secure and user-friendly platform for family members to preserve and share their culinary traditions. Includes a customizable recipe book with extensive options for personalization. Features a user-authentication system that allows users to create and manage their own recipe books. Includes custom artwork and a unique website design. Read more
Designed and developed the Crafting Comrades website as a shared hub for my high school friends' Minecraft worlds and modpacks. The website uses a database to store mod packs and worlds, and allows users to browse and download them. It also features a live server status of the running Minecraft server and a live map of the world. Read more
SpotifySongDisplay is a personal project that has evolved through multiple iterations, culminating in a polished, macOS-native application that visually enhances the Spotify listening experience. Starting as a basic Python desktop app, it transformed into a system-integrated top-bar display on macOS, showcasing album art, song titles, and dynamic color theming that seamlessly blends with the macOS user interface. Read more
Mini Tools
Developed a set of tools for SSH, SCP, and Git integration into Raycast. These tools allow users to easily connect to remote servers, transfer files, and manage Git repositories from the command line. Dynamically loads the user's SSH configuration for connection details and Git SSH accounts. Built a Git tool to set the current Git user config from Raycast and replace old Git user commits with new ones. These tools are designed with safeguards to prevent unintended or unwanted actions. Read more
Explored desktop application development with Tauri and Rust by creating a Bouncing Basketball experiment. Used Rust for programming physics interaction between the balls and the edges of the screen. This playful proof-of-concept was a fun way to learn Rust and Tauri. Read more
This project is a 3D water playground simulation. It uses the Three.js library to render the 3D graphics. The simulation includes a water surface that can be interacted with using the mouse. The water surface is implemented using a shader that simulates water physics. The project is implemented in JavaScript and is available on GitHub. Read more
Personal iOS app designed to provide students with a streamlined and at-a-glance view of their class schedules directly on their iPhones and iPads. Featuring both calendar and lock screen widgets, Simple Timeline offers immediate access to upcoming and current class information, helping students stay aware of their daily schedule and manage time more effectively. Read more
Education
Oregon State University in Corvallis, Oregon GPA: 3.9, Summa Cum Laude
Oregon State University in Corvallis, Oregon GPA: 3.8, Honor Roll for all terms
La Cañada High School in La Cañada Flintridge, California Took AP Computer Science Java & AP Computer Science Principles Recieved the La Cañada High School AP Computer Science Award
Experience
La Cañada High School in La Cañada Flintridge, California Part of the LC iTeam, a group of students who work with the school's technology department to help students and teachers. Developed, implemented and maintained school-wide bell schedule application for class times and class schedule planning. Collaborated with teachers and students to develop and improve the application to meet the needs of the school and students. go.lciteam.club
Oregon State University in Corvallis, Oregon Performed data analysis for the NANOGrav Collaboration under Dr. Xavier Siemens. Created data visualizations using Matplotlib and Seaborn to generate noise budgets and power-law spectra for pulsar timing data used in gravitational wave detection. Uncovered a filtering error in the preliminary processing of the 20-year dataset and created comparative visualizations against the validated 15-year dataset to illustrate the anomaly to the research team. Built performance profiling scripts to track memory usage and runtime metrics across thousands of iterations for HPC resource optimization.
Awards
Received the La Cañada High School AP Computer Science Award for excelling in the AP Computer Science Principles and AP Computer Science Java courses. The award was given or technical excellence, collaboration with classmates, and demonstrating great creativity.
Achieved Eagle Scout rank, from La Canada's Troop 507 in 2022. My Eagle Scout project installed new decking and sandbox features for the La Canada Child Education Center. The project included planning, design, fund raising and implementation. I used lidar scanning and 3D modeling to design the area, performed iterative materials choices based on fund raising success with constant communication and updates to the Child Education Center.
Skills
I have used HTML, CSS, JavaScript, and TypeScript for many web development projects including this website. I have experience with creating responsive and accessible web pages, web applications, and interactive simulations.
I have used Svelte, React, and Tailwind CSS for my web development projects. I have experience with creating and maintaining web applications, interactive data visualizations, and websites that communicate with APIs and databases.
I use Node.js and SQL / SQLite for a lot of my websites. I have many projects that use Node.js as server-side software for hosting. I also have hosted, managed, maintained, and integrated many databases into web applications and user interfaces.
Extensive experience with modern JavaScript and TypeScript across full-stack web development, from interactive front-end components to server-side APIs and type-safe application code.
I have used Java mostly for Minecraft mods and Minecraft server software plugins. I have also used Java for applications and larger projects. This is what I have the most experience with aside from web development. I have experience with JavaFX and Java Swing for GUI applications.
I have used Python extensively for machine learning and deep learning projects with PyTorch, NumPy, and TensorFlow, as well as for data analysis, web development, automation, and scripting.
I have used C++ for mostly school projects and have experience with making terminal applications and small games. My greatest C++ project is the Terminal Graphics Library.
I have used C extensively for systems programming, parallel computing with MPI and OpenMP, and low-level memory management in academic and research computing environments.
I have used Lua for a handful of advanced Minecraft ComputerCraft programs that can be used to monitor item storage and automation flow.
I have used Git for version control and collaboration on many projects. I have experience with Github Actions and CI/CD.
I have used Linux for my server operating system and have experience with managing and deploying services on Linux servers. I use Windows and MacOS for my desktop operating systems and have experience with development on both platforms.
I regularly use Blender for 3D modeling and animation and have experience with using it for creating assets for games and applications. I have used Figma and GIMP for graphic design and have experience with creating assets for web and applications. I have also used Figma for prototyping and designing a few applications.
I have used Docker to containerize applications and services. I have experience with deploying, managing, and creating Docker images and containers.
I have used VSCode as my primary code editor for many years and have experience with extensions and customizations.
I have used IntelliJ for Java development and have experience with the IDE and its extensions. I have also made a few plugins for IntelliJ.
I have a lot of experience with using the terminal and shell scripting. I have used the terminal for development, deployment, and server management and have written many shell scripts for automation.
Extensive experience building and training deep learning models with PyTorch, accelerated with CUDA on HPC clusters. Proficient in NumPy for scientific computing and data processing pipelines.
Experience developing real-time graphics applications with OpenGL, including GLSL shader programming for GPU-accelerated procedural generation, terrain rendering, and normals computation.