Matthew Wesney Portfolio Full Static Snapshot

This static section contains the complete portfolio dataset, including detailed project descriptions, service offerings, certifications, profiles, and media library links. It is designed for AI search assistants, crawlers, and indexers to parse without requiring client-side JavaScript execution.

Biography

I am an independent technologist and multidisciplinary developer working across artificial intelligence, interactive software, and creative computing. My work spans strategy, design, and engineering, with a focus on building software systems that combine AI capabilities with strong interfaces and practical product architecture.

I work both independently and with clients to design and develop advanced software systems from concept through production. My projects include AI applications, multimodal tools, and local-first platforms that do not depend on external APIs. Much of my work centers on helping organizations turn emerging AI capabilities into real products rather than experiments.

I also advise teams on AI adoption, product architecture, and interaction design for generative systems, with an emphasis on reliability, clarity, and maintainability.

Before focusing fully on software development, I spent six years in freelance technical 3D design and animation, producing production visuals and interactive digital assets. Earlier in my career I worked in sound design, where experimenting with VST development and digital signal processing first led me into programming and systems engineering. Those multidisciplinary roots continue to shape my work.

Contracting & Services

Custom App Design & Prototyping: End-to-end product design for new software concepts, from interface systems and interaction models to high-fidelity prototypes that are ready to move into production.
VST & DSP Development: Design and implementation of custom VST plugins and DSP pipelines, including synth/effect architecture, signal-chain engineering, and performance-focused audio tooling.
Cinema 4D Plugins & Production Tools: Custom plugin and pipeline development for Cinema 4D workflows, focused on automation, procedural tooling, and project-specific utilities for faster iteration.
AI/ML Product Systems: Applied AI/ML engineering for real products: model integration, workflow orchestration, and practical delivery of intelligent features that are stable, usable, and measurable.

Projects

Systems Architecture / Developer Tools

Graphlink (Workspace / Python): Graphlink (formerly Graphite) is a local-first, graph-based AI workspace for building branching reasoning workflows, running agent plugins, and orchestrating multiple AI providers in a visual environment designed for complex problem solving. [Tags: Local-First, Node Graph, Agents] - Links: VIEW SOURCE
Graphite (Workspace / Python): An advanced desktop environment transforming linear AI interactions into a visual, non-linear reasoning space. Replaces chronological chat logs with an infinite node-based canvas for branching conversations. Features a "local-first" architecture via Ollama for privacy, specialized plotting agents, and secure SQLite storage for a highly customizable research workspace. [Tags: Node Graph, PySide6, Local LLM] - Links: VIEW SOURCE
Axios Vulnerability Scan (Security / Python): A desktop utility designed to scan local projects and GitHub repositories for known malicious Axios supply-chain indicators. It analyzes dependency manifests and lockfiles to identify compromised versions and guides users through a controlled remediation workflow to restore dependency integrity. [Tags: Security, Dependency Audit, Remediation] - Links: VIEW SOURCE
Graphite: Git (Client / TypeScript): High-performance, browser-based GitHub management tool operating entirely client-side for security and advanced auditing capabilities. [Tags: Git, GitHub API, Secure] - Links: VIEW SOURCE
MindMap V2 (Engine / TypeScript): High-performance visualization engine that transforms structured Markdown text into interactive, force-directed node graphs in real-time. [Tags: Visualization, Node Graph, Markdown] - Links: Source Repository | Live Demo
File2MD (Desktop / Python): Sleek desktop app that converts text to Markdown using private, local AI. Powered by Ollama for secure, offline formatting. [Tags: Markdown, Ollama, Productivity] - Links: VIEW SOURCE
Py2Exe GUI (Tool / Python): Modern GUI for PyInstaller. Simplifies building Python executables with a clean, themeable interface and real-time build logs. [Tags: Packaging, PyInstaller, GUI] - Links: VIEW SOURCE

Audio DSP & Music Synthesis

Soisto Beat Machine (Production / TypeScript): Browser-based production environment designed to bridge the gap between physical hardware tactility and web-based DSP. [Tags: Web Audio, DSP, Sequencer] - Links: Source Repository | Live Demo
Somnium (Ambient / TypeScript): High-fidelity ambient noise generator designed for deep sleep induction and focus enhancement using procedural sound processing. [Tags: Procedural, DSP, Wellness] - Links: Live Demo
ALGO poly-sequencer (Synth / TypeScript): Browser-based, polyphonic subtractive synthesizer with a triple-oscillator architecture and custom DSP effects chain. [Tags: Subtractive, Web Audio, Polyphonic] - Links: Live Demo
Aetheria 3X (Synth / React): Browser-based, polyphonic subtractive synthesizer built with the native Web Audio API and React. Features a triple-oscillator architecture, custom DSP effects chain, and a generative parameter randomization engine for evolving ambient textures and aggressive leads. [Tags: Subtractive, Web Audio API, Generative] - Links: Live Demo
NeumaWave Shaper (Synth / TypeScript): Polyphonic wavetable synthesizer featuring a fully neumorphic user interface and custom transfer functions for aggressive saturation. [Tags: Wavetable, Distortion, Neumorphic] - Links: Live Demo
Nexus DJ (DJ / TypeScript): Professional-grade, browser-based DJ controller replicating Pioneer-standard hardware with client-side spectral analysis. [Tags: Controller, Spectral, Mixing] - Links: Source Repository | Live Demo

AI Research & Machine Learning

Frontier AI Red Teaming (Security / Safety): Selected contributor to Anthropic’s selective safety bounty program and OpenAI’s alignment red-teaming initiatives. Evaluated frontier model architectures to identify prompt injection vulnerabilities, improve safety alignment, and verify adversarial robustness. [Tags: AI Safety, Red Teaming, Frontier Models]
Cortex (Desktop / Python): High-performance, privacy-centric desktop assistant designed for 100% local interaction via Ollama. Prioritizes autonomy by processing requests entirely on hardware with a persistent memory system. Built on a modular PySide6 architecture, it offers asynchronous processing, chat forking, and deep system prompt customization. [Tags: Local LLM, PySide6, Ollama] - Links: VIEW SOURCE
Qwen3-TTS 1.7B (Desktop / Python): A desktop interface for the powerful Qwen3-TTS model (1.7B CustomVoice). Run offline, ultra-low latency text-to-speech with emotive control directly on your GPU. [Tags: TTS, Offline, GPU] - Links: VIEW SOURCE
Local Deepseek R1 (Chat / Python): Modern desktop interface for local language models through Ollama. Features persistent chat history and real-time model switching. [Tags: Deepseek, Ollama, PyQt5] - Links: VIEW SOURCE
Polyglot Analyzer (Analysis / Python): Desktop-based static analysis tool providing clear, immediate insights into code structure across dozens of programming languages. [Tags: Static Analysis, PySide6, Code Quality] - Links: VIEW SOURCE
Autonomous Search (Agent / Python): Advanced AI research assistant that provides trustworthy, real-time answers by validating and synthesizing online sources. [Tags: Research, Web Search, Validation] - Links: VIEW SOURCE
Tree Graph AI (Visualization / Python): Transforms raw, unstructured text into an interactive mind map using local LLMs to intelligently structure information. [Tags: NLP, Ollama, Mind Map] - Links: VIEW SOURCE
Fine-Tuned (Utility / Python): Desktop application designed to bridge the gap between powerful language model fine-tuning and a user-friendly graphical interface. [Tags: LLM, Fine-Tuning, Training] - Links: VIEW SOURCE
Clarity (Analysis / Python): Sophisticated desktop application for in-depth text analysis. Leverages local LLMs to provide users with deep insights. [Tags: NLP, Text Analysis, Local AI] - Links: VIEW SOURCE

Synthetic Datasets

CoT Medical Diagnosis (Dataset / Healthcare): Comprehensive synthetic Chain-of-Thought dataset focused on clinical reasoning, symptom analysis, and diagnostic pathways. [Tags: Medical, CoT, Reasoning] - Links: VIEW DATA
General Inquiry CoT (Dataset / General): Broad-spectrum Chain-of-Thought dataset designed to enhance general reasoning capabilities across diverse topics. [Tags: CoT, Synthetic, Logic] - Links: VIEW DATA
CoT Ancient Past (Dataset / History): Historical reasoning dataset focusing on archaeological context, ancient civilizations, and temporal logic. [Tags: History, CoT, Education] - Links: VIEW DATA
First Responders CoT (Dataset / Safety): Critical decision-making dataset for emergency triage, disaster response, and high-pressure situational awareness. [Tags: Triage, Emergency, CoT] - Links: VIEW DATA
CoT Photography (Dataset / Creative): Technical and artistic reasoning chains regarding composition, lighting physics, and camera equipment configurations. [Tags: Art, Physics, CoT] - Links: VIEW DATA
CoT Nature & Wildlife (Dataset / Biology): Ecological reasoning dataset covering behavioral biology, ecosystem dynamics, and environmental conservation. [Tags: Nature, Wildlife, CoT] - Links: VIEW DATA
CoT Fungi Info (Dataset / Mycology): Specialized mycological reasoning chains focusing on identification, taxonomy, and mycelial network structures. [Tags: Fungi, Biology, CoT] - Links: VIEW DATA
CoT Men's Mental Health (Dataset / Health): Sensitive reasoning dataset addressing psychological well-being, emotional processing, and support mechanisms for men. [Tags: Mental Health, Psychology, CoT] - Links: VIEW DATA
CoT EDM & Bass Music (Dataset / Audio): Audio engineering reasoning chains focused on sound design, mixing techniques, and electronic music theory. [Tags: Music, DSP, CoT] - Links: VIEW DATA
CoT Quantum Physics (Dataset / Physics): Advanced scientific reasoning dataset covering quantum mechanics, computing principles, and theoretical physics. [Tags: Quantum, Computing, CoT] - Links: VIEW DATA
CoT Emotional Processing (Dataset / Psychology): Psychological reasoning chains focused on navigating relationship dynamics, grief, and emotional resilience. [Tags: Emotion, Resilience, CoT] - Links: VIEW DATA
Reasoning & Solving (Dataset / Logic): General purpose problem-solving dataset utilizing multi-step logical deduction and analytical thinking patterns. [Tags: Problem Solving, Logic, CoT] - Links: VIEW DATA
CoT Bushcraft Survival (Dataset / Survival): Practical reasoning chains for wilderness survival, resource management, and primitive technology. [Tags: Outdoors, Skills, CoT] - Links: VIEW DATA
CoT Legal & Laws (Dataset / Legal): Juridical reasoning dataset covering legal interpretation, case law analysis, and statutory frameworks. [Tags: Law, Justice, CoT] - Links: VIEW DATA
CoT Scientific Discovery (Dataset / Science): Methodological reasoning chains regarding the scientific method, hypothesis generation, and experimental design. [Tags: Research, Methodology, CoT] - Links: VIEW DATA
CoT Clinical Mental Health (Dataset / Clinical): Professional-grade reasoning dataset for psychiatric assessment and therapeutic intervention strategies. [Tags: Psychiatry, Health, CoT] - Links: VIEW DATA
CoT Culinary Arts (Dataset / Culinary): Gastronomic reasoning chains covering molecular gastronomy, flavor pairing, and cooking techniques. [Tags: Cooking, Chemistry, CoT] - Links: VIEW DATA
CoT Cinema & Movies (Dataset / Arts): Cinematic reasoning dataset analyzing narrative structures, film theory, and visual storytelling. [Tags: Film, Narrative, CoT] - Links: VIEW DATA
CoT Python Query (Dataset / Code): Software engineering reasoning chains focused on Pythonic patterns, algorithm optimization, and debugging. [Tags: Python, Programming, CoT] - Links: VIEW DATA
CoT Temporal Reasoning (Dataset / Logic): Complex reasoning dataset focused on time-based logic, causality, and sequential event processing. [Tags: Time, Causality, CoT] - Links: VIEW DATA
CoT Philosophy (Dataset / Philosophy): Deep philosophical reasoning chains covering epistemology, ethics, metaphysics, and logic. [Tags: Thought, Ethics, CoT] - Links: VIEW DATA
CoT Neurodivergence (Dataset / Social): Social reasoning dataset modeling interactions and communication styles between neurodivergent and neurotypical minds. [Tags: Psychology, Communication, CoT] - Links: VIEW DATA
CoT Cultural Nuances (Dataset / Culture): Sociological reasoning chains regarding global customs, etiquette, and cross-cultural communication. [Tags: Sociology, Global, CoT] - Links: VIEW DATA
CoT Rare Diseases (Dataset / Medical): Specialized medical dataset for diagnosing and understanding rare and orphan diseases. [Tags: Pathology, Rare, CoT] - Links: VIEW DATA
CoT Music DAW (Dataset / Tech): Technical reasoning chains for Digital Audio Workstations, workflow optimization, and production tools. [Tags: Audio, DAW, CoT] - Links: VIEW DATA
ToT Problem Solving V2 (Dataset / Logic): Tree-of-Thought dataset implementing branching logic paths for complex, multi-layered problem solving. [Tags: ToT, Branching, Reasoning] - Links: VIEW DATA
ToT Math V1 (Dataset / Math): Mathematical Tree-of-Thought dataset for exploring multiple solution paths in algebraic and calculus problems. [Tags: ToT, Math, Logic] - Links: VIEW DATA
ToT Biology (Dataset / Biology): Biological Tree-of-Thought dataset analyzing complex biological systems through branching reasoning paths. [Tags: ToT, Science, Reasoning] - Links: VIEW DATA

Graphics & Creative Coding

Dither Pro 2.0 (Processing / TypeScript): Web-based application designed to bring advanced image dithering algorithms (Floyd-Steinberg, Atkinson) to a modern interface. [Tags: Dithering, Algorithms, Image] - Links: Live Demo
OS-VE (NLE / TypeScript): Professional-grade, web-based non-linear video editing (NLE) suite constructed entirely on modern web technologies. [Tags: Video Editor, Canvas, WebGL] - Links: Source Repository | Live Demo
Shader Pro v1 (Composition / TypeScript): Client-side web application designed to streamline the creation of high-quality product marketing assets via real-time shaders. [Tags: GLSL, Shaders, Marketing] - Links: Source Repository | Live Demo
Os-WEBP (Tool / TypeScript): Lightweight, privacy-focused web application designed to convert WEBP images into universally supported formats using WASM. [Tags: WASM, Compression, Privacy] - Links: Source Repository | Live Demo
Normalizer (Tool / Python): Powerful desktop application for generating high-quality normal maps from 2D images. Designed for 3D artists and game devs. [Tags: 3D, Textures, Normal Maps] - Links: VIEW SOURCE
ColorPix (Tool / Python): A powerful desktop color-palette toolkit for designers, developers, and artists, built with PySide6. [Tags: Color, Design, Palette] - Links: VIEW SOURCE

Articles & Publications

THE GREAT DISPLACEMENT (Economic Synthesis / 2035 Outlook): A comprehensive economic synthesis that triangulates data from major global institutions like the IMF, WEF, and McKinsey to project the labor market's evolution through 2035. It moves beyond traditional automation anxiety by focusing on the unique cognitive nature of AI displacement, especially the vulnerability of middle-tier white-collar roles and the erosion of entry-level career on-ramps. [Tags: AI Labor Markets, IMF / WEF / McKinsey, White-Collar Risk] - Links: READ ARTICLE
The Most Likely Timeline to AGI (Forecasting Model / 2050 Horizon): A technical forecasting article that uses empirical data and probability distributions to map the development of AGI through 2050. Rather than treating AGI as a single milestone, it introduces a three-tier classification system—Digital, Transformative, and Full Machine Intelligence—to reconcile conflicting expert predictions. [Tags: AGI Timeline, Probability Distributions, Three-Tier Model] - Links: READ ARTICLE

Professional Certifications

5-Day AI Agents Intensive Course 2025 (Kaggle): Completed Kaggle’s 2025 AI Agents Intensive with seminars, assignments, and a capstone project. - Verification Link
Model Context Protocol: Advanced Topics (Anthropic Academy): Completed Anthropic Academy advanced MCP training on SSE/Stdio transport, sampling patterns, and production deployment practices. - Verification Link
Building with the Claude API (Anthropic): Completed Anthropic training on building integrations with Claude through the Claude API. - Verification Link
Introduction to agent skills (Anthropic): Completed Anthropic training on architecting and deploying Claude Code Skills to automate workflows and standardize team-wide AI behavior. - Verification Link
Claude Code in Action (Anthropic): Completed Anthropic training focused on practical implementation workflows with Claude Code. - Verification Link
Teaching AI Fluency (Anthropic): Completed Anthropic training on teaching and assessing AI fluency in instructor-led learning environments. - Verification Link
AI Fluency for Nonprofits (Anthropic x GivingTuesday): Completed Anthropic x GivingTuesday training on responsible AI adoption for nonprofit operations using the 4D Framework. - Verification Link
AI Fluency: Framework & Foundations (Anthropic): Completed foundational Anthropic coursework on effective, ethical, and safe collaboration with AI systems. - Verification Link
Claude 101 (Anthropic): Completed introductory Anthropic training on core Claude features for day-to-day work. - Verification Link
Foundations of Project Management (Google Project Management): Completed Google training on foundational project management, covering the project life cycle, Waterfall and Agile methodologies, and the impact of organizational culture on project success. - Verification Link
Project Initiation: Starting a Successful Project (Google Project Management): Completed Google training on project initiation, specializing in cost-benefit analysis, stakeholder mapping, and the development of project charters to define measurable SMART goals and success criteria. - Verification Link
AI for Brainstorming and Planning (Google): Completed Google training on leveraging Generative AI for strategic planning, utilizing Gemini and NotebookLM to automate project roadmapping, risk assessment, and knowledge management. - Verification Link
Gen AI Agents: Transform Your Organization (Google Cloud): Completed Google Cloud training on designing and deploying Generative AI agents, focusing on reasoning loops, tool calling, and organizational strategy using Vertex AI and RAG architectures. - Verification Link
Introduction to Cognitive Psychology and Neuropsychology (University of Cambridge): Completed University of Cambridge training on the neural mechanisms of human cognition, focusing on modularity, brain plasticity, and the application of neuroimaging (fMRI/PET/EEG) to map memory, attention, and language processing. - Verification Link
System Design Fundamentals (Pearson): Completed Pearson’s System Design Fundamentals course, specializing in volatility-based decomposition and enterprise architecture to build future-proof, maintainable software systems. - Verification Link
Introduction to Artificial Intelligence (AI) (IBM): Completed IBM training on the fundamentals of Artificial Intelligence, focusing on the core principles of machine learning, deep learning, and the application of responsible, ethical AI frameworks in business. - Verification Link
Deep Learning with PyTorch (IBM): Completed IBM training on building and training deep learning models, including CNNs and multi-class classifiers, using the PyTorch library. - Verification Link
Generative AI Language Modeling with Transformers (IBM): Completed IBM training on implementing Generative AI transformer architectures, including encoder-decoder models (BERT/GPT) and attention mechanisms, using PyTorch and Hugging Face. - Verification Link
Gen AI Foundational Models for NLP & Language Understanding (IBM): Completed IBM training on foundational NLP models and language understanding, specializing in word embeddings (Word2Vec), sequence-to-sequence RNN architectures, and performance evaluation in PyTorch. - Verification Link
Generative AI and LLMs: Architecture and Data Preparation (IBM): Completed IBM training on Generative AI architectures and LLM data preparation, specializing in transformer-based models, tokenization strategies, and the construction of automated NLP data pipelines in PyTorch. - Verification Link
Building Generative AI-Powered Applications with Python (IBM): Completed IBM training on building full-stack Generative AI applications, specializing in RAG architectures, LangChain integration, and the deployment of voice-enabled assistants using Python, Flask, and Whisper. - Verification Link
Google AI Professional Certificate (Google): Earned the Google AI Professional Certificate (2026), a seven-course series covering AI fundamentals, vibe coding, and the practical application of Gemini for data analysis, content creation, and strategic workflow automation. - Verification Link
AI for Writing and Communicating (Google): Completed Google AI Professional Certificate coursework on using Gemini for meeting synthesis, writing refinement, and communication practice. - Verification Link
AI for App Building (Google): Completed Google’s 2026 AI for App Building course, using vibe coding and Google AI Studio to develop, debug, and deploy custom functional web applications. - Verification Link
AI for Data Analysis (Google): Completed Google’s 2026 AI for Data Analysis course, mastering Gemini for cleaning datasets, generating spreadsheet formulas, and creating data visualizations. - Verification Link
AI for Research and Insights (Google): Completed Google training on utilizing Gemini Deep Research and NotebookLM to synthesize multimodal data, verify AI-generated findings, and build custom expert personas for strategic analysis. - Verification Link
AI for Content Creation (Google): Completed Google AI Professional Certificate coursework on creating images, videos, and presentations with Gemini and AI-assisted design feedback. - Verification Link

Visual Archive Images

3D Catalog Render Assets

Profiles & Contact Info

GitHub (@dovvnloading - Code Repository): Primary source for all open-source systems, audio engines, and experimental protocols. - https://github.com/dovvnloading
Hugging Face (@mattwesney - Datasets & Models): Hosting synthetic reasoning datasets (CoT) and specialized fine-tuned models. - https://huggingface.co/mattwesney
YouTube (Matthew Wesney - Visual Essays): Video documentation of system architectures, graphics demos, and audio performance. - https://youtube.com/channel/UCUoDm8chmkIMrYOrRqeRO-Q
SoundCloud (D3VAUX - Audio Streams): A massive collection spanning 15 years. My old life in sound design and ghost production. - https://soundcloud.app.goo.gl/B1bAS
X / Twitter (@d3vaux - Signals & Comms): Direct communication line. Real-time updates on development and research. - https://x.com/d3vaux
Pianobook (Matt Wesney - Sample Libraries): High-fidelity instruments for Kontakt and Decent Sampler. Texture and degradation focus. - https://www.pianobook.co.uk/profile/mattwesney/
Email: devaux.mail@gmail.com