Technology executive
I turn AI strategy into working systems. Over 20 years I've led technology across healthcare, ecommerce, PE-backed growth, and a national nonprofit federation, building global engineering organizations that modernize core platforms and take AI from pilots to production.
I'm Vice President of Software Engineering at YMCA of the USA, where I lead platform strategy, AI-driven engineering, and a blended engineering organization across continents, building the national products local YMCAs rely on. Before that, I led technology at a PE-backed firm, reporting to the CEO, and led ecommerce engineering at Spirit Airlines.
In my roles, I've partnered with CEOs, executive teams, and boards: tying technology investment to outcomes they can measure, building blended global teams that own what they ship, and making AI part of how work gets done rather than a side project. Over my career, I've led engineering teams across North America, Asia, Latin America, and Europe. I also build AI projects on my own time through Melitrus, my independent lab, to stay close to building hands-on.
MBA, Indiana University Kelley School of Business
MS Computer Science, Illinois Institute of Technology
BTech Information Technology, Kurukshetra University
Generative AI for Business Leaders, Indiana University
A record of building the teams, platforms, and capabilities organizations run on.
YMCA of the USA
Lead platform strategy and AI adoption across engineering for the national products local YMCAs rely on: building an AI-driven development model, partnering with leading AI technology companies, making build, buy, and partner decisions across the platform portfolio, and modernizing the platforms the national network depends on.
Walker Sands
As the senior-most technology leader reporting to the CEO, owned due diligence, integration architecture, and platform consolidation across two acquisitions as the firm scaled to $23M.
Walker Sands
Led a FinancialForce PSA/ERP implementation integrated with Salesforce, improving project scoping efficiency by 25%, lifting staff utilization by about 10%, and raising profitability by about 8%.
Spirit Airlines
Led web, mobile, and loyalty engineering through a complete re-architecture of Spirit.com and a full mobile rebuild with zero revenue disruption, and built ecommerce optimization from scratch, driving an 18% conversion lift.
Vein Clinics of America
As Director of BI & Enterprise Applications for a 63-site healthcare system, built analytics and data science capabilities and unified EMR, CRM, and clinical data in an enterprise data warehouse.
The MX Group
Led application development across a portfolio of B2B clients.
PCNation
Built a price intelligence tool that gave the business clear visibility into competitor pricing.
Advance Innovative Solutions
Built a data-mining solution for claim science and recovery, and white-label mobile apps for prescription refills and pharmacy information.
AI helped me build most of Veruti quickly. Deciding what it should do with ambiguous documents was the part only people could do, and designing that judgment to compound was the most valuable part of the build.
A practitioner's comparison of ADLC5 and GitHub's Spec Kit, two spec-driven delivery workflows for AI coding tools, and what enterprise teams actually need from them.
These are personal projects I build or support on my own time, separate from my role at YMCA of the USA. I'm not paid for any of them, and my own projects generate no revenue.
Melitrus
In developmentAd testing usually happens after the money is spent. AdDNA moves it before launch. It predicts how a piece of creative will perform, so teams can fix the weak ad, back the strong one, and stop paying to learn what didn't work. It gives marketing teams evidence, not gut feel, before a campaign goes live.
Get in touch about AdDNAMelitrus
In developmentMost AI pulls data out of documents. Veruti catches what's wrong in them: inconsistencies, ambiguity, and data that looks right but isn't. Every answer is traced to its root in the source, turning messy documents into decision-ready, reasoning-backed data. When it isn't sure, it asks a person, then builds that answer into a persistent memory for future decisions. Industry-agnostic by design, starting with mortgage lending.
Get in touch about VerutiPro bono advisor
LiveEvery caregiver handoff starts with the same question: when did the baby last eat? LETO Baby answers it before anyone asks, with feeds, sleep, and diapers on one shared timeline. And because hands are usually full, it logs by voice with Alexa. Everyone caring for the baby stays in sync, without a single group text.
Visit LETO Baby ↗Contributor
Open sourceAI coding agents write code faster than teams can trust it. ADLC5 holds them to account: a deterministic kernel, not the agent, decides when work is done, and only on current evidence. SOUL, its reasoning discipline, keeps agents from wandering into confident tangents, and the process scales with each change's risk. I contribute with a focus on enterprise adoption.
View on GitHub (ADLC5) ↗For speaking, or to compare notes on AI in the enterprise.