# AI Hub and your commercetools setup

Understand how AI Hub reuses your existing commercetools Project setup for discovery and checkout.

AI Hub does not introduce a separate commerce system. It connects AI providers and checkout protocols to the same commercetools APIs your storefront already uses.

## What you need in your Project

AI Hub relies on the following commercetools resources and settings:

- Product data from [Products](/api/projects/products.md) and [Product Variants](/api/projects/products.md#productvariant)
- A dedicated [Store](/api/projects/stores.md) for each agentic channel context
- Optional [Product Selections](/api/projects/product-selections.md) assigned to a Store to scope the available assortment
- Optional [Product Tailoring](/api/projects/product-tailoring.md) in the Store context for tailored titles, descriptions, images, and attributes
- Optional Distribution Channels and Supply Channels assigned to the Store for price and inventory resolution
- [Shipping Methods](/api/projects/shippingMethods.md) that are available for the Store context
- Pricing, tax, and promotion logic applied during [Cart](/api/projects/carts.md) and [Order](/api/projects/orders.md) processing
- Order lifecycle management after checkout completion

The same APIs remain the source of truth. AI Hub adds channel integrations on top of them.

For the rationale behind dedicated Stores, Product Selections, Product Tailoring, Distribution Channels, Supply Channels, and Shipping Methods, see:

- [Introduction to agentic commerce](/learning-prepare-for-agentic-commerce/prepare-for-agentic-commerce/introduction-to-agentic-commerce.md).
- [Product catalog readiness](/learning-prepare-for-agentic-commerce/prepare-for-agentic-commerce/product-catalog-readiness.md).

Learn more about how to prepare your Project for agentic commerce in the self-paced Prepare your Project for agentic commerce standalone learning module.

## Discovery flow and catalog exposure

During discovery, AI Hub prepares and publishes catalog data to provider-specific formats.

In most implementations, the flow follows these steps:

1. Create a Store that represents the agentic channel context.
2. Assign [Product Selections](/api/projects/product-selections.md) to that Store to scope the available assortment.
3. Use [Product Tailoring](/api/projects/product-tailoring.md) in that Store when the channel needs tailored Product data.
4. Publish price and availability from the Store context using the configured Distribution Channels, Supply Channels, and [Inventory](/api/projects/inventory.md) data.
5. Map Product and Product Variant fields to provider-specific feed attributes.
6. AI Hub keeps updates in sync through scheduled jobs, event-driven updates, or both, depending on the integration.

For detailed setup guidance, see:

- [Introduction to agentic commerce](/learning-prepare-for-agentic-commerce/prepare-for-agentic-commerce/introduction-to-agentic-commerce.md) for Project, Store, and Channel setup patterns.
- [Product catalog readiness](/learning-prepare-for-agentic-commerce/prepare-for-agentic-commerce/product-catalog-readiness.md) for Product Selection strategy, Product Tailoring, product data requirements, and feed synchronization concepts.

## Checkout flow and protocol behavior

During checkout, AI Hub receives protocol requests from an AI channel and translates them to commercetools API operations.

For protocols such as the Universal Commerce Protocol (UCP) and other channel-specific flows, the flow generally includes:

1. Create or update a Cart from buyer and Line Item input.
2. Resolve shipping details and recalculate totals.
3. Authorize payment through the configured Payment Service Provider.
4. Create an Order in commercetools after successful authorization.
5. Send Order lifecycle updates back to the AI channel when required.

All validations continue to come from your existing configuration, such as inventory constraints, pricing rules, and promotion applicability.

For the rationale behind checkout sequencing, extension points, and payment authorization flows, see:

- [Checkout readiness](/learning-prepare-for-agentic-commerce/prepare-for-agentic-commerce/checkout-readiness.md).
- [Payment readiness](/learning-prepare-for-agentic-commerce/prepare-for-agentic-commerce/payment-readiness.md).

## Key setup decisions for merchants

To avoid configuration issues, agree on these decisions before launch:

- Which Store keys represent each AI channel entry point.
- Whether Product Selections should be assigned to each Store.
- Which Products need Store-specific Product Tailoring, if you use Product Tailoring.
- Which Distribution and Supply [Channels](/api/projects/channels.md) should be used for pricing and availability.
- Which [Shipping Methods](/api/projects/shippingMethods.md) should be associated with each Store context.
- Which locales, currencies, and countries are used for feed and checkout contexts.
- Which API Client scopes are required for discovery and checkout components.

Before production rollout, validate your configuration and test flows using the [Production readiness checklist](/learning-prepare-for-agentic-commerce/prepare-for-agentic-commerce/production-readiness.md).

## What AI Hub does not do

- AI Hub does not replace your existing commerce logic.
- AI Hub does not maintain a separate source of truth for inventory.
- AI Hub does not bypass Store and Product Selection constraints.

## Related pages

- For OpenAI data mapping details, see [Products data mapping for OpenAI](/agentic-commerce/ai-hub/openai/product-data.md).
- For Google feed and checkout details, see [Google module overview](/agentic-commerce/ai-hub/google/overview.md).
- For Store configuration details, see [Stores](/api/projects/stores.md).
- For catalog scoping details, see [Product Selections](/api/projects/product-selections.md).
- For channel and stock details, see [Channels](/api/projects/channels.md) and [Inventory](/api/projects/inventory.md).

## Related pages

- [Area overview page with navigation](/agentic-commerce.md)
- [Previous page: Overview](/agentic-commerce/ai-hub/overview.md)
- [Next page: Manage agentic channels](/agentic-commerce/ai-hub/agentic-channels.md)