🌙
A circular profile photo of Matthew Wesney

Matthew Robert Wesney

AI Developer | Creative Technologist | Full-Stack Engineer

About Me

A dedicated and relentlessly driven developer with a deep passion for building innovative AI and machine learning systems. My creative journey spans over 16 years — from professional music and sound design to 3D art, animation, and now, advanced AI development.

My approach to mastering development is intensely hands-on; I believe the best way to learn is to build. This philosophy has guided me through an intensive period of creating a wide range of applications, each serving as a practical exploration of new technologies. This process is driven by a singular purpose: to build tools that think, help, and inspire. My background in the arts gives me an intuitive edge in designing these systems to feel alive — blending creativity with precision.

My Approach

My development philosophy is rooted in deep, first-principles thinking. I thrive on deconstructing complex problems to their core components and rebuilding them into elegant, intuitive systems. This method allows me to spot intricate patterns others might miss and architect solutions from unconventional, highly effective angles.

I excel in focused environments that value deep work, transforming a simple concept into a fully-realized, production-ready application. While my passion is for the focused act of creation, I believe the best systems are born from clear communication and mutual respect.

Services

AI & Application Development

  • Custom AI desktop applications (Python, PySide6)
  • Ollama-integrated tools for local LLM inference
  • Fine-tuning and dataset preparation
  • Autonomous agents and memory systems
  • Multi-agent communication frameworks

UI/UX & Product Design

  • Clean, modern desktop interface design
  • Minimal, functional, and accessible UIs
  • Wireframing and concept visualization
  • 3D visualization to enhance user feedback

Creative & Multimedia

  • Sound design and synthesis for software
  • Generative ambient soundscapes
  • 3D rendering, animation, and visualization
  • AI-assisted music and audio tools

Featured Projects

Graphite

Visual LLM Workspace

A desktop application that transforms linear chat into a dynamic, non-linear visual workspace using local LLMs.

View Project

PrismXL

AI Image Generation GUI

A high-performance desktop GUI for generating images with diffusion models, featuring advanced controls and a responsive UI.

View Project

Cortex

Desktop AI Assistant

A secure, modular desktop application for local LLM interaction, built with a production-ready architecture.

View Project

Tree-Graph-MindMap

Interactive Knowledge Visualizer

Transforms unstructured text into dynamic, explorable mind maps, aiding researchers in structuring complex information.

View Project

Autonomous Web Search

Verified Research Agent

A real-time web research agent with automatic source validation, providing trustworthy, verified answers.

View Project

Fine-Tuned

LLM Fine-Tuning GUI

A user-friendly interface for rapid LLM experimentation and deployment, democratizing model customization.

View Project

normalizer

3D Normal Map Generator

A fast, intuitive desktop tool for converting 2D images to production-quality normal maps for 3D artists.

View Project

RL-Tic-Tac-Toe

Dueling DDQN Agent

An educational tool combining PyTorch and PySide6 for real-time reinforcement learning and visualization.

View Project

Extended Project Showcase

AI Assistants & Search

LLM / Ollama Utilities

Datasets & Training Resources

Beyond applications, a significant part of my work involves engineering high-quality data, specializing in synthetic datasets for advanced reasoning and Chain-of-Thought (CoT) processes.

Many of these are built using my custom dataset generator, "Genesis," and undergo a rigorous cleaning and formatting pipeline post-generation. The goal is to provide clean, structured, and potent data for fine-tuning models capable of complex problem-solving.

Reasoning_Problem_Solving_Dataset

151k Q&A pairs (STEM, applied reasoning)

ToT_Reasoning_Problem_Solving_V2

72.1k structured Q&A

General_Inquiry_Thinking-Chain-Of-Thought

6k Q&A (broad reasoning)

ToT-Math-V1

26.8k pairs (math problem-solving)

ToT-Biology

23k pairs (biology reasoning)

CoT_Music_Production_DAW

7k pairs (music production & AI)

A Lifelong Creator

Born in Columbus, Ohio, in 1989, I’ve spent my life creating across mediums — sound, visuals, and now code. Before development, I spent 16 years as a musician and sound designer, creating sample packs that shaped the early dubstep scene, and even free VSTs. Some of these plugins were listed among the Top 10 Best Free VSTs of All Time and were downloaded +50k times worldwide.

That same creative instinct led me into game development, where I first engineered AI enemy behavior systems in Unreal Engine. Over time, that curiosity evolved into a full transition toward AI and machine learning — where I now focus my energy on building intelligent, open-source tools that expand human capability.

Why Work Together

Portfolio-Driven

An extensive public project history demonstrates capability without relying on credentials alone.

End-to-End Ownership

From concept through deployment, the full development lifecycle is managed independently and meticulously.

Versatile Expertise

Equally comfortable with AI infrastructure, UI/UX, backend systems, and specialized creative domains.

Production Mindset

Every project ships with clean code, clear documentation, and a user-focused experience at the forefront.