I build and ship production AI systems. End to end, fast, on a team or on my own.
RAG and agent systems, tool-use loops, voice-AI pipelines, evals, and the payments and infra around them. Three live AI products shipped in the last year, solo, from architecture decision to deploy. 20-plus years building regulated SaaS, and I have led engineering teams and consolidated platforms serving millions.
A coaching operating system: one platform to build, deliver, and grow a coaching practice, replacing the five-plus disconnected apps coaches juggle for calendars, payments, programs, and messaging. An AI program builder turns a plain-language brief into a full curriculum, and a Claude agent runs a real tool-use loop against 16 tools that read and write live data. Stripe Connect handles payouts, scheduling runs on RFC 5545 RRULE, and it ships across eight services. I built all of it, and real coaches and clients transact through it.
AI-powered interview prep for anyone getting ready to interview, in any industry. You practice out loud with an AI interviewer that talks back: real conversation, not multiple choice. Every answer gets scored on structure, specificity, and impact against a 0-100 readiness model. It mines your best STAR and CARL stories through guided conversation, then feeds you weakness-weighted drills. It runs on a real-time voice pipeline (Pipecat, Daily.co WebRTC, Deepgram, Silero VAD, GPT-4.1, ElevenLabs) with GPT-4o handling the scoring.
Real-time scoring, ticketing, and fundraising for judged competitions. Multiple judges score live, synced across audio, video, and scores on a single clock, with configurable scoring scales that adapt to any judged event. An AI pipeline transcribes judge audio and writes performance and event summaries with Claude, and Stripe Connect splits payouts to event hosts. I built it end to end.
I work across the whole AI stack: agents and tool-use loops, RAG, voice pipelines, and evals, plus the frontend, data layer, payments, and deploy around them.
I drop into an existing codebase and team cleanly: your stack, your reviews, your standards. I review code, mentor engineers, and I am as comfortable owning one feature as a whole product. I also move fast with AI as a force multiplier, driving Claude Code to ship evaluated, production systems.
I work directly with the people who own the problem. I scope the build with them, explain the tradeoffs in plain language, and ship something real, not a demo. I can read a codebase and tell you what the risk and the remediation will cost.
I have led engineering teams through real change: a company sale, the consolidation of acquired products serving 1.5 million-plus users, and a legacy rebuild that cut delivery time roughly 5x. I have sat on both sides of an acquisition, and I am still the one in the code.
Stack
What I'm building with today.
TypeScript
JavaScript
Python
React
Next.js
Vite
Tailwind CSS
Node
Fastify
FastAPI
PostgreSQL
Prisma
ChromaDB
Anthropic
Claude Agent SDK
OpenAI
Deepgram
ElevenLabs
Pipecat
Daily.co
Ollama
Stripe / Stripe Connect
Payload CMS
Docker
Railway
Vercel
AWS
Sentry
Trigger.dev
Playwright
Claude Code
VS Code
Previously
C#
.NET
ASP.NET
SQL Server
SSIS
SSRS
Angular
DataDog
New Relic
SCORM
xAPI
Who I am
I build, and I have led at scale.
I started my career in the U.S. Air Force, including a deployment to the Middle East. When I got out, I went after a degree in physical therapy, but the insurance side of healthcare pushed me toward technology instead. I became a network admin, then a developer, then a senior developer. Nearly every role since has been in EdTech, eLearning, or coaching platforms in regulated industries. That depth of focus is deliberate.
One of my early career-defining roles was at HealthStream, where I was part of the team that rewrote their core platform. When I joined, the stock was trading under a dollar. Today they are a publicly traded healthcare SaaS company doing over $300 million a year. I saw firsthand what a solid platform rewrite can do for a company's trajectory.
In 2008 I joined SimplyDigi as an early technical leader with significant equity. I was core to the platform's growth from the ground up, scaling it past 600,000 users before we sold to Vector Solutions. At Vector I led the consolidation of six acquired platforms into a unified SaaS architecture serving over 1.5 million users. At eLuma I rebuilt a legacy K-12 teletherapy platform and roughly 5x'd delivery velocity.
These days I build. In the last year I shipped three live AI products, mostly solo and mostly AI-generated code, from the agent systems and voice AI down through payments and deploy. I have run engineering at scale, and I still write the code.
I have been fully remote since 2008, now based in Tallahassee, Florida, running distributed teams before everyone else had to figure it out. Every company I have worked with has been in a regulated environment: K-12 EdTech, healthcare-adjacent systems, coaching platforms handling sensitive data. I know what it takes to build systems that survive compliance scrutiny.
