Taproot
Mt. Joy
softball
AI agents
I'm an M.S. Information Systems student at the University of Cincinnati and a builder who can't sit still. I ship real software solo by directing AI agents, then I go figure out how to get it in front of people, because a great product nobody sees isn't worth much.
I'm endlessly curious, and I'd rather be a beginner at something new than coast on what I already know. That's how I went from sales and consulting to shipping a live SaaS, an App Store app, and my own internal tools in a year.
Off the keyboard, you'll find me at a show or a baseball game, on the softball field, somewhere on a trail, or conjuring up my next adventure.
I'm AI-native at everything, not just code. Research, content, outbound, ops, building, I run all of it through AI. It's not a tool I reach for, it's how I work.
I work at the systems level. I don't write much code by hand. I direct AI agents and design the systems around them, evals before specs, specs before code, plus reusable skills, so the output is reliable instead of vibes.
Building is the easy half. The longer I build, the more I think figuring out who it's actually for, where they are, and how it reaches them matters just as much as the product or the problem it solves. I've started designing for distribution from day one instead of bolting it on at the end.
I don't just use AI tools, I build the machine around them. I've assembled 70+ custom Claude Code skills, MCP servers, subagents, and hooks into a system that runs my whole operation: research, content, code review, outbound, and deployment. A few favorites:
Memory for your AI. Taproot gives Claude, ChatGPT, and any other assistant a long-term memory: it brings the right context back up when you need it, and connects the dots across everything you have saved.
AI tools forget everything the moment a session ends, so you re-explain yourself over and over. Taproot makes your context stick, and resurface on its own.
Solo, end to end: a Mac helper app (Swift, Sparkle auto-update, Apple-notarized) that syncs your Obsidian vault to an encrypted cloud mirror, an MCP server that exposes it to any AI client, Stripe billing, and a full onboarding flow.
Envelope encryption, a stack of security hardening across nine launch-blocker PRs, and a strict ship-vs-deploy discipline, all backed by 375+ tests.
A cold-outbound engine I built end to end to source, enrich, verify, and personalize B2B leads for my own ventures.
The small, local businesses I wanted to reach barely show up in the big databases like Apollo and Clay. I needed a way to find and contact the people those tools miss.
A pipeline that discovers businesses from niche directories and the open web, enriches them free-first before any paid lookup, writes AI-personalized sequences, and scores every lead.
A web dashboard and CLI to run it all, Instantly.ai send and reply analytics, and a cost-per-lead audit layer that surfaces the real economics of every run.
A consumer iOS hydration app built around a Tamagotchi-style virtual pet that degrades when you ignore it and perks up when you drink.
Most hydration apps are earnest and forgettable. I built one with a personality people actually want to keep alive.
Designed and built the full app solo: the game loop, paywall, onboarding, and App Store optimization, then submitted it to the store.
My first iOS app. I'd never shipped to the App Store, so I kept the scope small on purpose: take one simple idea, build it end to end, and actually get it out the door, learning Swift and SwiftUI from scratch.
A college baseball app that puts live D1 scores, schedules, lineups, box scores, Top 25 rankings, and transfer-portal activity in one place, wrapped in a retro scoreboard look.
College baseball has never been more popular, but the apps fans use are an afterthought: scores buried in clunky interfaces, missing lineups, Top 25 lists that barely name the teams.
The full app end to end (React Native and Expo): live scores with auto-refresh, game detail with inning-by-inning line scores and box scores, a date-browsable schedule, Top 25 and RPI rankings, and the transfer portal, all on live ESPN and NCAA data.
A typed API layer over several sports-data sources with smart caching (fast refresh for live games, longer for finals), plus a cohesive retro design system with light and dark mode.
Co-architected an Explainable AI Coach with a team, delivering bias-tested cash-flow recommendations for SMBs inside a governance framework for data privacy, fairness testing, and compliance.
Ran market research on CPG trends and Braun's competitive positioning to map untapped product categories into a market-expansion roadmap.
Provided project-management support for audit readiness and tax due diligence on a 60+ entity student-housing REIT roll-up, and mapped financial process flows into SOPs for a national non-profit.
Streamlined customer order and service-request management across phone, email, and EDI, sourced product solutions with suppliers, and contributed to a cross-functional wallet-share project.
Built data-driven B2B marketing campaigns and delivered client-facing software demos, translating technical product features into business value.
Built logistics and on-time delivery reports and ran a container-optimization project.
When I'm not building, I'm usually moving. I lift, play tennis, softball, and golf, and in winter I'm snowboarding or skiing. I'm at Cleveland Guardians games whenever I can get there, and I love live music, indie rock especially (Mt. Joy, flipturn, Palace, and Peach Pit are on heavy rotation).
The rest of the time I'm chasing an adventure. I plan trips around getting outside: last year I went to Costa Rica to surf, this year I hiked through Guatemala, and there's always a next one in the works. Curiosity is sort of the whole theme, on and off the keyboard.