Founder · Builder · Domain-First

I go inside
industries. Then I
build the fix.

I don't research problems from the outside. I work inside them. I've processed PI cases at a law firm, watched healthcare teams waste 45 minutes on paperwork that should take seconds, and seen a shipping market rupture in real time. Then I built the AI to fix each one.

→ Worked inside a PI law firm   · building CaseZero
→ Saw healthcare prior auth waste   · built Sanctr
→ Watched Etsy block Pirate Ship   · built LogixVast
→ Promised my sons a game   · built Bible Heroes Run
4Products Built
$53BNext Market
9Yrs Enterprise (USPS HQ)
3Languages

Live Status

Currently Building

Four products. All moving.

Coming Soon
CaseZero
Right now: Lien resolution engine in design. First 3 PI firms in outreach. Pre-seed thesis written. $53B market, zero connective AI layer.
Early Access
Sanctr
Right now: Onboarding orthopedic practices for a 30-day free pilot. Epic FHIR sandbox connected, real-patient testing underway.
Seeking First 20
LogixVast
Right now: Pursuing Etsy Approved Partner status. Warm leads on Reddit. Seed extension conversations underway.
Live on App Store
Bible Heroes Run
Right now: Android version in progress. Sunday School classroom mode scoped. A promise to my sons — kept.

The Work

Products & Ventures

Each one started with living inside the problem first.

Coming Soon
Legal AI · Personal Injury · Lien Resolution
CaseZero
The connective intelligence layer plaintiff-side law doesn't have yet.

Personal injury law generates $53B in attorney fees annually. Tools exist to automate each stage in isolation — intake, medical processing, demand, negotiation, closeout, archive. But nobody has built the intelligence layer that connects all six, compounds learning across every case, and does what insurance companies can already do and plaintiff firms cannot: predict what a case is worth before a single demand letter is sent.

The wedge is lien resolution — a manual, legally complex, 4-hour process on every single settled case. I'm building it from the inside: I currently work at a personal injury law firm, processing cases, tracking liens, drafting demand letters, negotiating with adjusters every day. The dataset I'm building cannot be replicated by a competitor who hasn't lived it.

Lien Resolution PI Counsel AI Drafting Suite Python AWS RBAC / Audit Trail
eCommerce · Logistics · SaaS
LogixVast
The shipping software that doesn't punish growth.

Etsy blocked Pirate Ship in October 2024 with less than a week's notice. 5M+ sellers were trapped between a product they loved and a monopoly that gouges them. I built the third option: multi-carrier rate shopping, Etsy + Amazon + Shopify sync, transparent pricing, and human support with a 2-hour SLA.

Next.js Django PostgreSQL Redis Stripe AWS
Healthcare AI · Prior Auth
Sanctr
Prior auth appeals. Automated. 45 minutes → 3 seconds.

30% of orthopedic prior auth requests are denied on first submission. Writing each appeal takes 45 minutes — and 1 in 5 denied claims are never appealed at all. Sanctr generates a complete, clinically grounded appeal letter in under 3 seconds. Formatted, editable, ready to fax. HIPAA-eligible on AWS, with Epic FHIR integration for real patient and coverage data.

AI/LLM Epic FHIR HIPAA-eligible AWS Python React
iOS Game · Kids · Faith
Bible Heroes Run
A promise kept.

I had no mobile dev experience. I learned to code because I promised my sons — Joshua and Caleb — that I would build them a Bible game. The result: a Mario-style iOS platformer covering Genesis to Revelation, 1,000 levels, 10 playable Bible heroes each with scripture-rooted superpowers, co-op mode, original MIDI music, 100% free with no ads ever. Built with Vite, Vanilla JS, HTML5 Canvas, Capacitor iOS. Now live on the App Store.

This is the one that shows who I am, not just what I can build. That's the most important project.

Vite Vanilla JS HTML5 Canvas Capacitor iOS

Enterprise Credibility

Enterprise Architecture

9 years inside USPS Headquarters — the systems thinking behind every product above.

Live Demo
USPS ML Governance Framework

A working reference architecture for how a federal-scale organization should govern machine learning — built to show SA-level ML fluency, USPS-specific architecture judgment, and production-proven patterns. Interactive, single-file, fully self-contained.

ML Governance SA vs ML Engineer Ownership Agentic AI Frameworks

The Pattern

About

I spent 9 years at USPS Headquarters — product management, innovation, and payment programs across systems serving 600K+ employees. That work taught me what scale actually means, what compliance actually requires, and what it costs when you get it wrong.

I'm not an architect moonlighting as a builder. I'm a domain-first founder — I go inside an industry, live the exact pain firsthand, then build the AI to fix it. I worked inside a PI law firm and I'm building CaseZero. I watched healthcare teams burn time on paperwork and I built Sanctr. I watched a shipping market rupture in real time and I built LogixVast. And I promised my sons a Bible game, so I built Bible Heroes Run.

The enterprise background makes me build things that actually scale and survive. The domain-first building keeps me sharp, hungry, and close to the real problem. Each one makes the other better.

⚙️

Enterprise AI Architecture

ML platforms, MLOps, compliance-first design — at USPS scale across AWS, Azure, and GCP.

🚢

Full-Stack Builder

Next.js, Django, React, Python, Capacitor iOS — idea to shipped product, solo.

🔒

Federal Compliance & Security

NIST 800-53, FISMA, HIPAA — built on a federal data governance playbook I helped create.

👨‍👦

Dad First

My sons Joshua and Caleb have a game built for them, by me, from scratch.