

The Short Version
My name is Lance. I've spent the last fifteen years in academia, teaching English across three continents, designing assessment systems for global audiences, coordinating programs, and teaching courses ranging from project management to digital literacy to innovation and entrepreneurship. I worked as a banker before that, managed teams throughout, and learned that precision and structure transfer across domains even when the tools change. Now I am applying my experience to AI systems architecture, building document intelligence systems from first principles and documenting every step of the process.


MagVin Labs exists because I had a problem worth solving. Fifteen years of documents accumulated across three continents, sitting idle because they could not be searched, compared, or analyzed. Nothing I found did exactly what I needed, so I decided to build it myself.
I also had a conviction: that personal data should remain under personal control. Cloud services offer convenience and scale, but some information belongs closer to home. MagVin Labs is built on the premise that individuals should have the option to run their own intelligence systems, not because the cloud is wrong, but because choice matters.
The blog documents that process: the ambition, the overreach, the corrections, and the frameworks that emerged.
Why MagVin Labs Exists


What I Bring
I am not a coder by training. I am a systems thinker who learned to work with code because the problems I wanted to solve required it. What I bring is two decades of structured problem-solving, including curriculum architecture, assessment design, program coordination, and team leadership, now applied to AI systems.
The technical skills are developing. The judgment was already there.


MagVin Labs is built with AI assistance. I use Claude, GPT, and Gemini as co-architects, thinking partners for design decisions, code generation, and documentation. The coding happens through Claude Code and Google Antigravity.
This is not a workaround for missing skills. It is the skillset itself: orchestrating AI systems to produce working architecture. The ability to collaborate effectively with AI tools is exactly what building AI systems now requires.
As the Lab matures, more components will run on local models, moving toward the sovereignty the project is designed to demonstrate. The current architecture uses cloud AI to build toward local AI. The irony is not lost on me. It is the path.
How I Build
What's Next
MagVin Labs remains under active development. The Project Log documents the first generation of document intelligence systems on the MagVin Desk (v0-v5). The Lab series documents what came next: containerized architecture, local model integration, and the infrastructure for true data sovereignty.
The goal is a personal AI lab I fully control, secure, searchable, and capable of surfacing connections across decades of accumulated information. That is the destination. The Project Log documents the journey.
If you are interested in the work, the thinking, or the person behind it, I would be glad to hear from you.

