Solutions engineer & AI automation consultant. I design, ship, and operate conversational AI in production — and the platforms around it. Every number here is from a live system I run today.
The craft. Architecture, data models, pipelines, interfaces, and the unglamorous operations layer — uptime, cost, guardrails. Built solo, owned end-to-end.
The proof. Four client platforms live in production. An AI agent on duty 24/7. A teaching cohort running the same patterns. Nothing on this page is a mockup.
When I arrived there was no mobile or automation function — just me, and a fleet running on manual process. So I built the department, and architected the shipboard infrastructure end-to-end.
Up to ~10,000 devices on a single large ship, across all three brands. Strategic satellite deployments, a device staging lab in Doral stood up with the Apollo Group, and the training + handover docs so each ship’s System Managers and Fleet Assistants could run it without me.
CBP facial-recognition onboarding at the gangway. Guest-facing apps through the full software lifecycle — and I was the final review gate before anything shipped to the App Store, iOS and Android both. Modernized culinary, restaurants and housekeeping operations. Security sign-off on everything guest-facing.
From one person to a team of four. A fleet-wide mobile backbone where there had been none — and a human checkpoint on every release, years before I’d apply that same instinct to AI.
SME for automating & integrating custom software into Zebra hardware. Led nationwide rollouts — The Home Depot (US + Canada), Lowe's, Target, Dick's, Pilot Travel, Woolworths Australia.
High-level federal security architecture spanning several agencies on CISA's CDM mission. Under NDA — specifics stay sealed; the scope did not.
Where it started — modernized enterprise mobility & endpoint security across the org. The grounding in systems that simply have to stay up.
I was there on day one. That was the aha — the moment my brain started to imagine again. Eighteen years of enterprise plumbing suddenly had a new verb: talk to it. I haven't stopped building since.
Forty years ago, Star Trek gave us a computer you simply talk to.I'm finally building it.
I've been a Trekkie my whole life. Conversational AI isn't a new idea — it's an old promise, arriving. We aren't inventing it. We're shipping it.
A national accounting conference needed attendee and sponsor commerce in one place — registration, payments, sponsor packages, contracts — on a date that could not move.
One SPA doing real work: Stripe checkout for attendees and sponsors, sponsor contracts generated as PDFs on the fly, stored behind private signed URLs, passwordless sponsor auth. I consolidated 32 serverless functions down to 9 to fit platform limits, and delivered through repeated live UAT rounds with the client.
It's outgrowing the word "portal." The foundation for an Event OS is already in — an event-app shell with content management and issue tracking, feature-flagged and in preview. The next layer is conversational: an assistant attendees and sponsors can simply talk to — schedule, logistics, contracts, support — the same production conversational-AI pattern I run elsewhere, pointed at a live event.
Done-for-you AI automation for small CPA firms. Editorial design system, fixed-price, ownership by default — plus the full schema.org / llms.txt layer most sites skip.
momandpop.ai →
The teaching arm — a live cohort where operators build the agent systems I run. Same type stack, inverted palette. The members' edition of the same world.
nocodesociety.com →
I've already lived the second half of the conversational-AI problem — the part after the demo. Multi-channel agents real people talk to daily. Memory that stays accurate over months. Cost controls, permission tiers, audit trails, and the judgment to put a human checkpoint exactly where the model shouldn't be trusted alone.
I work where it actually counts: messy real-world data, compliance-minded clients, systems that have to be right — not just impressive. Eighteen years of enterprise rigor, now pointed at AI. And because I teach this, I bring a team along, not just ship beside them.
antonio@momandpop.ai