Use the Smart Data Modeler to define how your products are structured for AI consumption.
AgenticLift recommends a single Product Type to structure all product data for AI platforms. A flat, consistent schema makes it easier for agents to reason across products, simplifies feed generation, and reduces data mapping complexity. Multiple Product Types are supported if your catalog requires it, but each additional Product Type increases the complexity of the data import and JSONata mapping configuration.
The Product Type must be created before data ingestion.
Generate the Product Type
Use the Smart Data Modeler with agentic channels to analyze a sample of your catalog, map it to the attributes your target platforms require, and generate a platform-ready Product Type. For the full workflow, including how to review the proposal and import or transform your catalog, see Smart Data Modeler - Agentic channels.
The Smart Data Modeler supports OpenAI and Stripe Agentic Commerce Suite. When you generate the proposal, select All of the above if you plan to support both platforms, to avoid catalog changes later. Google Gemini channels don't use the Smart Data Modeler; you configure them during agentic channel setup.
Mandatory attributes
Each AI platform defines its own mandatory attributes. A missing mandatory attribute excludes the product entirely from the feed rather than merely disadvantaging it. For the attributes common across platforms, see Mandatory attributes.
Product Selection
Create a Product Selection to define which products are included in the Universal AI Feed, then assign it to the Store. Business users can use this Product Selection to toggle individual products on or off for the agentic channel.
Each SKU synced to a platform counts as one SKU-platform unit. For example, syncing the same SKU to both OpenAI and Gemini counts as two units. Before importing, verify that your catalog size multiplied by the number of target platforms does not exceed your contracted allowance.