The AI Builder Podcast
How Shopify Is Building for Agent-First Commerce with Mani Fazeli
Shopify VP Product Mani Fazeli explains why products must serve both humans and AI agents—from machine-readable catalogs and SimGym to agentic checkout.
What you’ll learn
- Products now operate across two simultaneous interfaces: a human experience layer and an agent decision layer.
- AI agents evaluate products through structured attributes, constraints and machine-readable data rather than visual merchandising alone.
- Shopify's SimGym uses simulated AI buyers to test purchasing decisions before ideas reach real customers.
- UCP and embedded checkout turn checkout into programmable infrastructure that lets agents transact without bypassing merchant logic.
- Agent-first commerce compresses the traditional funnel into find, evaluate and execute, making decision clarity and outcome completion the new conversion system.
About this episode
Commerce is shifting from a world where people browse, compare and buy directly to one where AI agents increasingly help evaluate products and execute transactions. Shopify VP Product Mani Fazeli explains why this changes more than the interface: every commerce product must now perform across a human experience layer and an agent decision layer at the same time.
Fazeli walks through how Shopify is rebuilding the foundations for both layers. Product data has to be legible to models, protocols such as UCP and embedded checkout have to let agents transact without bypassing merchant logic, and SimGym uses AI buyers to simulate purchasing decisions before launch. Checkout does not disappear when the interface does; it becomes programmable infrastructure that enforces price, inventory and business rules.
The leadership implication is a new operating model for commerce. Funnels compress from discovery, evaluation and conversion into find, evaluate and execute. Teams need machine-readable catalogs, reliable transaction protocols and simulation before launch—while preserving the emotion, trust and brand experience that still shape human preference.
