Skip to content
View severian42's full-sized avatar

Highlights

  • Pro

Block or report severian42

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Maximum 250 characters. Please don't include any personal information such as legal names or email addresses. Markdown supported. This note will be visible to only you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
severian42/README.md

Typing SVG

Profile Views LinkedIn Hugging Face Website Spotify Email


About

I'm a Forward Deployed AI Engineer at Adobe and an independent AI researcher with a background that doesn't fit neatly into any single lane — audio engineering, ecological restoration, fine woodworking, music production, and machine learning all inform how I build and think.

My work sits at the intersection of applied LLM systems and original framework research. On the enterprise side, I architect and deploy AI solutions at scale. On the research side, I'm developing original mathematical machinery — SEFA (Symbolic Emergence Field Analysis), CMT (Complexity Magnitude Transform), and SRL (Symbolic Resolution Ladder) — and applying them to open problems in dynamical systems and information geometry.

I believe nature is the longest-running engineering testbed in existence. Most of my best ideas came from paying attention to it.


What I Build

Enterprise / Production

  • AI-native application stacks for enterprise clients, with a focus on retrieval architecture, agent orchestration, and LLM deployment pipelines
  • Forward deployment work at Adobe, bridging the gap between model capability and real-world product integration

Open Source Tools

  • GraphRAG-Local-UI — The most complete local GraphRAG implementation with full UI, indexing, query, and visualization. 2.3k stars.
  • Vodalus Expert LLM Forge — End-to-end LLM fine-tuning pipeline using MLX and Unsloth with RAG-grounded dataset generation
  • SIREN — Field-theoretic approach to unbounded memory in large language models
  • Firewatch — TypeScript monitoring and alerting system

Research Frameworks

Models & Datasets on Hugging Face

  • ANIMA — LLM fine-tuned on biomimicry and STEM examples, trained to reason through biological analogy
  • Nexus — Dataset encoding interconnected knowledge graph structures

Tech Stack

Languages

Python TypeScript JavaScript Next.js

ML / AI Frameworks

PyTorch MLX LangChain Ollama HuggingFace

Infrastructure & Tools

FastAPI Docker Jupyter Anthropic OpenAI


Education

Degree / Certificate Institution
B.S. Computer Science Full Sail University
B.S. Audio Engineering Full Sail University
A.S. Ecology & Computer Technology Ventura College
A.I. Machine Learning & Data Science Zero To Mastery Academy
Prompt Engineering Bootcamp Zero To Mastery Academy

GitHub Stats

GitHub Streak

GitHub Trophies


Activity Graph


Building things that matter at the edges of what's possible — reach out if you want to collaborate.

Pinned Loading

  1. Symbolic-Emergence-Field-Analysis Symbolic-Emergence-Field-Analysis Public

    SEFA is an information-geometric framework for detecting emergent, symbolic structure in continuous or spectral fields. It acts as a symbolic lens that amplifies coherent, interpretable patterns by…

    Python 1

  2. Computational-Model-for-Symbolic-Representations Computational-Model-for-Symbolic-Representations Public

    Glyphs, acting as collaboratively defined symbols linking related concepts, add a layer of multidimensional semantic richness to user-AI interactions by serving as contextual anchors that guide the…

    56 4

  3. GraphRAG-Local-UI GraphRAG-Local-UI Public

    GraphRAG using Local LLMs - Features robust API and multiple apps for Indexing/Prompt Tuning/Query/Chat/Visualizing/Etc. This is meant to be the ultimate GraphRAG/KG local LLM app.

    Python 2.3k 292

  4. Vodalus-Expert-LLM-Forge Vodalus-Expert-LLM-Forge Public

    Dataset Crafting w/ RAG/Wikipedia ground truth and Efficient Fine-Tuning Using MLX and Unsloth. Includes configurable dataset annotation editor Gradio UI.

    Jupyter Notebook 194 24

  5. SIREN SIREN Public

    A Field-Theoretic Approach to Unbounded Memory in Large Language Models

    Python 20

  6. Firewatch Firewatch Public

    TypeScript 10 2