Unlock Opportunities at Cross Border Summit: November 3 - 5, 2026

Login Signup Learn More
GFA 491 June 16, 2026 49:20

Inside Build My Online Store and the Rise of Agentic Commerce

Download Subscribe

Show Notes

Mike Michelini sits down with Christian Garcia, CEO of Paraliving and a nearly five-year member of the GFA team, for an unfiltered look at how a small e-commerce operation is going all-in on AI to run leaner, smarter, and with full ownership of its own data. From building a custom project management app that centralizes every data point in the business, to experimenting with agentic commerce through BMOS (Build My Online Store), Christian and Mike share what's actually working, what's still being figured out, and why sellers who aren't getting more technical right now are falling behind.

Episode Timeline

  • From Spreadsheets to Systems: Christian traces his path from aircraft engineering and a pandemic pivot into e-commerce, through building massive Excel dashboards and Looker Studio setups, to eventually creating a fully custom web app that centralizes all business data for AI to read and act on.

  • Why Off-the-Shelf SaaS Won't Cut It:: No single platform solves everything. Christian argues that custom-built tools, where you own your own data,  give sellers a bigger long-term advantage than stitching together third-party subscriptions that hold your data hostage.

  • Landed COGS and Logistics Automation: A deep look at how Christian built a system to auto-calculate landed cost across sea and air shipments,  down to FIFO inventory tracking by pulling supplier packing lists, shipping costs, and customs data into one place.

  • A Visuale Knowledge Base: Christian reveals how he's scraping top Amazon listing images to train AI on what great content looks like, then using it to generate listings, A+ pages, and storefronts automatically.
  • Agentic Commerce and BMOS: Mike explains Build My Online Store: a product-feed layer (not a storefront) that makes seller catalog data readable by AI agents using ACP and UCP standards, so buyers can transact without ever loading a website.

  • Practical Action Steps for Sellers: Christian closes with four takeaways: audit your current operations for blockages, get more technical, start saving your own data somewhere AI can freely access it, and keep forecasting what's coming next without overcomplicating things.

Topics Covered

Data as Your Biggest Business Asset Every piece of operational data — sales, logistics, cases, listings, supplier info — lives in someone else's platform right now. Christian's argument: pull it out, own it, and make it accessible to AI so it can surface real insights, trends, and recommendations without you having to ask.

Building vs. Buying: The Custom Software Case Rather than stitching together ten SaaS tools, Christian built one project management app that holds everything. The payoff is an AI agent that actually knows the business — current stock, sales velocity, campaigns, team approvals — and can reason across all of it in real time.

Why Amazon Sellers Need to Get Technical Mike and Christian both push back on the idea that sellers can outsource their way through the AI transition. The sellers who thrive will be the ones who understand what's happening technically — even if they're not writing the code themselves.

Agentic Commerce: The 1996 Moment Mike draws the analogy to the early internet. ACP (Agentic Commerce Protocol), championed by Stripe, and UCP (Universal Commerce Protocol), backed by Google and Shopify, are competing standards for how AI agents read and transact with e-commerce stores. BMOS is the team's bet on where this lands.

R&D Culture: Build Fast, Break Things, Move On Both Mike and Christian have abandoned more apps than they've shipped — and they're fine with that. Their advice: make something small, even something you'll throw away. The learning compounds, and in the AI era, starting over is cheap.

People / Companies / Resources Mentioned in this Episode

Download Options