nickhageman.com

Nick Hageman

engineer @ appleAI/ML | Robotics | AR/VR

Hey there!

I'm Nick, a software engineer based in San Diego 🌴

Now
SWE @ Apple
Previously
Apple Intern Β· Holo Reality Lab Research Assistant Β· John Deere Intern Β· UIowa Teaching Assistant

Some notable things I've done:

Projects

A collection of some school & hobby projects

🧠 NeuralNick

A macOS SwiftUI app + Python/LeRobot backend for an SO-101 robot arm: teleoperation, data collection, and on-device training and inference of ACT and VLA (SmolVLA, Ο€0.5) policies. It ends with me eating oatmeal by steering the arm with only an EEG headset.

RoboticsLeRobotPyTorchSwiftUIVLAEEG / BCI

nickOS

Created a bare-metal operating system from scratch running on an iPod Classic. Features a cooperative task scheduler, interrupt-driven input and audio, FAT32 filesystem, Pong, music playback/visualization, and a GBA emulator.

COperating SystemsEmbedded SystemsHardware Drivers

SmartDart

Secured $5,000 in funding for SmartDart: A computer-vision driven steel-tip dart system that combines accurate, automated scoring with dynamic solo gameplay and real-time feedback.

UIowa Innovation ChallengeTransfer LearningPythonOpenCVDeepDarts

Handheld Retro Game Controller

Designed and constructed a handheld gaming device leveraging the capabilities of the ESP8266 module and a Raspberry Pi. The choice of game for our device was Pac-Man, a classic arcade game known for its straightforward yet challenging gameplay.

ECE:3360 Embedded SystemsArduinoC++websocketsRaspberry PiJavaScript
🧠 Neural RGB-D Encoding screenshot 1

🧠 Neural RGB-D Encoding

Transmitting 3D data can be expensive, especially on hardware limited devices. Our approach to this problem was to create an end-to-end neural network sandwiched around an image codec for our encoding scheme.

PyTorchApplied MLPython

SlopeStats

πŸ‚ SlopeStats is a watchOS + iOS app which has multiple modes of tracking activity for skiing & snowboarding. It offers run tracking (speed, heart rate, altitude), Speed Mode (Ghost racing), Resort & Weather Info (API)

SwiftSwiftUICoreDataCoreMotionHealthKit

🌽 FarmVision

Modeled agricultural field data in Virtual Reality by utilizing John Deere Precision Ag APIs. Awarded "Best Data Collection Hack" at HackUIowa 2023.

HackUIowa 2023UnityC#Quest 2Python

HoloKinect

Developed visionOS UI for a realistic 3D video communication application on the Apple Vision Pro. (Mentored by Prof Tyler Bell and PhD candidate Stephen Siemonsma)

Holo Reality LabSwiftvisionOS

MoodTube

Fine-tuned a sentiment analysis model on YouTube comments. Created a chrome extension which did inference and visualized results by modifying the YouTube page's HTML.

ECE:5995 LLMsNLPFine-TuningBERTLoRA

LeetGPT

Developed a Chrome extension tool that provides LeetCode users with solutions to coding problems. Leveraged OpenAI's ChatGPT Language Model API for generative solutions.

OpenAI APIReactJavaScriptChromium
🌑️ IoT Thermometer screenshot 1

🌑️ IoT Thermometer

Created an IoT thermometer that communicates with a web server to provide the user with temperature values no later than 300 seconds ago. Withstanded water and drop tests. Conformed to strict design requirements.

ArduinoSerial CommunicationPython

CageVision

Created a spatial application which utilized an MMA API to gather upcoming event information to be displayed in a visionOS window. Conceptualized viewing MMA matches in augmented reality by introducing a 3D model of an octagon placed on a flat surface.

SwiftVisionOSXcodeRealityKit
🧠 Unsupervised RGB-D Scene Categorization screenshot 1

🧠 Unsupervised RGB-D Scene Categorization

Clustered scenes from a short film using RGB-D data. Included the depth information for clustering as we thought it may enhance the scene representation in regards to spatial relationships.

ECE:5995 Applied MLPytorchPythonPandasNumpy

πŸ’Ž Shards of the Grid

Our team developed a multiplayer game with Generative AI components in the form of a SaaS application using the Rails web framework.

ECE:5820: SELTOpenAI APIRubyHTML/CSS

πŸš— DriveSense

We utilized computer vision to monitor and record driver distractedness. We used a Raspberry Pi and Teachable Machine for our machine learning model. We also constructed a React dashboard application for data visualization.

ECE:5550: Internet of ThingsPythonOpenCVFirebaseJavaScript

The Gauntlet

Created a multi-stage game comprised of various challenges for our Virtual & Augmented Reality final project. Some of the challenges explored the mechanics of zero gravity, propulsion, enemy collision detection, animations, and AI navigation.

ECE:5995 VARUnityC#

πŸ“šENGR:2730 Computers in Engineering

Added features to Asteroids using topics including: OOP, dynamic memory allocation, SFML library, composition, inheritance, and polymorphism. Assisted 300+ students in learning advanced C++ concepts.

Teaching AssistantC++SFMLCMAKE
Generative AI Content Pipeline screenshot 1

Generative AI Content Pipeline

Used multiple Generative AI Tools to automate the content generation process for platforms such as YouTube & TikTok.

ECE:5995 Generative AI ToolsPythonOpenAI API

🐺 Werewolves

Created a multiplayer game using purely Java. Inolved multithreading, networking, and database management. Was chosen as a top team to present to the class.

ECE:3330: Software DesignJavaSQL