For 27 years I built learning experiences — inside institutions, then through my own ventures. In 2025 I launched Design X Factor. A year later, I quit. That word is on this page on purpose: reinvention starts with saying it plainly.
Instead of writing a new resume, I went into the lab. Every project below is framed the same way — the question I asked, what I built, and what actually happened. Including the ones that didn't work.
years in learning & technology
ventures launched — one I quit, and named openly
projects in eighteen months, solo
books published on Amazon
The domain I come from — and the first thing I rebuilt.
What does an enterprise LMS look like when it's built by one person who actually cares about the learning experience?
A multi-role learning platform — admin, instructor, and student dashboards — SCORM-compliant, with Stripe-gated courses, AI course and quiz generation, and QR-encoded credentials. Deployed on the Cloudflare edge.
Live and running for real courses.
Can a machine turn a PDF into a finished course — video, narration, quizzes, SCORM package — end to end?
An AI course factory: avatar video with voice cloning, five text-to-speech engines, AI imagery, interactive components, and SCORM 2004 export — with its own control center and editor.
The engine works and produces real courses. It's not a platform you can sign up for — it's my in-house production line, and my own research keeps pointing to it as a durable advantage: course production as a service, in English and Spanish.
Could I out-build the exam-prep market — for exactly one student?
All seven HESI A2 subjects, adaptive diagnostics that build a personal study plan, spaced repetition, interactive chemistry visualizations, and an instructor dashboard.
The student was my daughter. She reviewed every build at the dinner table, which is a harsher QA process than any client contract. She passed.
Build the architecture once. Let it serve very different humans.
Can a devotional app feel like a worn book instead of a PDF in a wrapper?
iOS and Android from one codebase — AI-illustrated series, a custom AI narration pipeline, reflection journals, three reading skins, offline-first.
Live on both platforms.
Can the same proven architecture serve a completely different human — a teen girl in a heavy moment — bilingual and free?
A non-clinical mental-wellness app for teen girls: one 90-second Daily Spark a day, a Fog Room for grounding exercises, everything on-device with no account required.
In build, reusing the Words for the Day backend. The decision log of what to reuse versus rebuild is the real lesson — that's platform thinking, not app building.
Can an app hold a patient's hand through the loneliest stretch of recovery — the weeks between surgery and the first follow-up?
An orthopedic recovery companion for knee-replacement patients: day-by-day guidance on what to expect, what to do, and when something is actually wrong — in the voice of a real orthopedic PA, designed accessibility-first for someone who is 70, in pain, and has one hand free.
In build. The clinical content is being reviewed by a practicing orthopedic PA before it reaches a single patient.
Can an AI book-writing companion run 100% on your phone — provable in airplane mode?
The architecture and product plan for an on-device authoring companion. Your knowledge never leaves the device — that's the entire wedge.
Before building the app, I'm running a go/no-go quality spike. If on-device drafting fails the test, the project dies. Kill criteria defined up front.
Tools I built because I wanted them to exist.
Can an AI tutor see your screen and teach by drawing on it — arrows, worked steps, equations?
The architecture for a desktop shell that orchestrates the AI subscription you already own plus free local voice — no API bills, no servers, sold once.
Early design. The architecture is complete; the product is not. Honest status: not shipped yet.
Can AI extract a book out of a person instead of writing it for them?
Interview-driven authoring: voice profiles, AI-conducted interviews, knowledge extraction, and synthesis into prose that traces back to what you actually said. Fully bilingual.
A working tool — it feeds the publishing line below.
How much SaaS can one person replace in a weekend?
My own meeting-notes capture, a personal tracking system, a CRM scaffold, and an e-signature tool for my own contracts.
I stopped paying for software I could build.
The boldest experiment: a GPU in my office as a production line.
Can a local AI pipeline produce print-ready books at near-zero marginal cost?
A full production pipeline — generation, quality control, 4× upscaling, print-ready PDFs — scripted end to end. Ten books produced, nine published on Amazon.
They sold approximately zero. Diagnosis: about 95% a traffic problem, not a product problem. The pivot: digital bundles on Etsy, Pinterest as the traffic muscle, two titles retired. The pipeline works; the lesson was distribution.
Can AI extract books out of me — in two languages?
Six titles on Amazon — self-help, faith, and fiction — written with my own offline tools. A seventh, an illustrated Spanish book on independent judgment in leadership, is in production.
All published and available.
Proof it translates: real businesses run on this.
Can a concierge wellness practice run its whole business on software it owns?
A public site, a client portal, and a custom CRM — scheduling, treatment records, outreach campaigns — with zero SaaS subscriptions.
The practice runs on it every day.
Every project starts as a falsifiable question, and the risky ones start with a go/no-go spike. If the answer is no, the project dies before it costs a year.
One proven architecture — edge backend, one mobile codebase — redeployed across products. Brilla stands on Words for the Day's shoulders. Speed is a system, not a sprint.
I work solo, but I don't work alone. AI drafts, illustrates, narrates, and codes with me — and a GPU in my office runs as a 24/7 production line.
I quit a company. I retired books that didn't sell. I parked tracks the evidence didn't support. And I wrote all of it down — that's what makes the wins believable.
If something here made you want to build — a platform, an app, a course engine, a strange idea that needs a first version — that's what my studio does.
Visit Xperience Labs