Hunter Williams

CompSci, IT Graduate

It's me. Hunter. I'm at the beach at Wassaw Island after kayaking two hours to get away from everyone after my skills.

Uh oh, you found me!

My name is Hunter. I make mods for video games and play with machine learning models.

Educational Background

I attended and completed a degree in computer science from Georgia Southern University at the end of 2024, and have an Associate of Applied Science in Information Technology from Southeastern Technical College I picked up along the way. These programs equipped me with both theoretical and practical expertise in computer science and IT, and probably caused all these gray hairs to suddenly start popping out. Dear god, I'm still in my 30's - please have mercy.

My Skills:

I'm pretty good in the following areas:

Analyzing complex problems and assessing the feasibility of implementing a variety machine learning algorithms to address them effectively.
Developing and managing software/web applications, including both frontend and backend systems.
Tracking down the strangest issues in your netcode, or straining my eyes over various debugging tools to similar ends.
Picking up new skills and technologies and adapting them to use professionally.
Server infrastructure deployment and management with Windows and Unix based systems and hypervisors.
Reverse engineering application code from headers and bytecode, so I can build my own sandcastle on top of yours.
Mastering music. Seriously, though I wouldn't want to do it as my job.

Meaningful words:

Here's some words instead of sentences. If you like any of them I can probably work for or with you.

Java (+Maven, Gradle)
Python
C# (.NET, MVC)
Rust (It's pretty.)

Machine Learning Application
Natural Language Processing
Dataset Cleaning
Data Analysis and Modeling

Server Administration
Database Design
API Development
Version Control (Git)

Cloud Computing
DevOps Practices
Web Development
Agile (Scrum) and/or XP

Check out some of my projects below! Find more at my GitHub.

Projects

Godot ELIZA

A modern implementation of the 1965 ELIZA chatbot by J.W., recreated using the Godot Game Engine and Python. This project highlights my ability to translate legacy code into contemporary frameworks, demonstrating technical proficiency and creative problem-solving.

Key Technologies and Skills

Godot Game Engine: Core functionality and interactive UI elements.
Python: Translated and adapted from the original ELIZA code.
Language Interoperability: Bridging Godot and Python seamlessly to ensure smooth integration.
UI Design: An intuitive and user-friendly interface.

Explore the releases on GitHub for more details and updates, or see a live implementation of ELIZA's code here.

The Dino-Store!

A playful yet functional e-commerce platform showcasing the power of modern web technologies. The Dino-Store is designed to deliver dinosaurs directly to your door while demonstrating dynamic client-side rendering using DOM diffing algorithms and a REST API-driven architecture.

Key Technologies and Skills

JavaScript: Core language for building dynamic and responsive web interactions.
Web Frontend/Backend Development: A seamless experience from user interface to server-side logic.
REST API Development: Designed and integrated RESTful APIs for data management and real-time updates.
Client-Side Dynamic Web Applications: Leveraged DOM diffing for efficient rendering and improved performance.

Explore the live site here, or view the codebase on GitHub.

A clip of the presentation for Lockstep, a machine learning algorithm for social bot detection.

Lockstep: An Ensemble Method

Lockstep leverages a combination of layered traditional machine learning models and graph neural networks to classify user behavior on social media, identifying patterns that resemble known bot activity. Designed for scalability, efficiency, and interpretability, Lockstep's architecture ensures that decisions made by the models remain transparent and cost-effective, making it a practical solution for large-scale behavioral analysis. By balancing simplicity with advanced analytical capabilities, Lockstep provides actionable insights while maintaining a low processing footprint per user.

Key Technologies and Skills

Python: Using Jupyter as a platform to analyze data.
Machine Learning: Hyper-tuning and application of random forests and embedded models.
Neural Networks: Leveraging graph neural networks to spot patterns in relationships.
Real World Application: Behavioral analysis based on recent data and designed to be generalized to modern platforms.

Find my work on this project here on github.

A definitely not AI-generated picture of... something plugged into an object similar to an electrical outlet, gushing water as if converting the electricity into it. Heh.

Python Firehose Processor

Designed to capture important information from a Bluesky Firehose relay, for use in machine learning research.

Key Technologies and Skills

Python: Uses standard python libraries.
Data Mining: Converts the real-time interactions from any of Bluesky's firehose relays into parquet format containing only useful information.
Multithreading: Uses multithreading to simultaneously listen to the bluesky firehose via websocket, record information, and provides chunk aggregation/pre-processing services for bulk-data assignments - all at once.

This project is both functional and available freely here on github.

Contact Me

Im on LinkedIn!

Connect with me on LinkedIn to learn more about my professional journey and stay updated on my work.

Email Me

Feel free to reach out for job opportunities, collaborations, or inquiries at:

[email protected]

Please note: No advertisements.

Check Out My GitHub!

Explore my projects and code on GitHhub account. If you’re searching for something specific and don’t see it there, don’t hesitate to send me an email!

Other News

Visit Again Soon

Thank you for visiting! Be sure to check back in the future for new updates and exciting content.