# AI Hub overview Learn about AI Hub, a centralized platform for managing agentic commerce through GenAI channels. AI Hub is a commercetools add-on product that extends the unified commercetools offering with agentic commerce through AI channels. It provides pre-built modules that connect your existing commercetools backend to popular Large Language Model (LLM) providers. This allows you to make your products discoverable and purchasable directly within these channels. ## Introduction to agentic commerce Agentic commerce is a major shift in how customers shop online. GenAI tools guide customers from product discovery to checkout within a single conversation. In traditional online commerce, shoppers manually navigate and browse for products through a storefront. Agentic commerce instead relies on AI to interpret the shopper's intent. This includes suggesting relevant products, making comparisons, and facilitating purchases. For shoppers, it provides a convenient and personalized shopping experience. For businesses, it opens a new channel for engagement and revenue generation. In an agentic commerce channel, an AI agent handles discovery, comparison, and decision-making on the customer's behalf within a conversational interface. This does not replace your storefront or your existing systems. It adds a new channel through which customers can reach and transact with your catalog, running on the same infrastructure you already rely on. Your existing commerce logic for pricing, tax, promotions, and inventory applies exactly as it does today. Learn more about how to prepare your Project for agentic commerce in the self-paced standalone learning module. ## How AI Hub works AI Hub connects three components: the LLM layer, the protocol layer, and the PSP layer. ### The LLM layer The LLM powers the agent's reasoning. It interprets what the customer wants, asks clarifying questions, compares products, and decides when to act. AI Hub connects the LLM providers that power these interactions for your Project. Supported providers are Google Gemini and OpenAI ChatGPT. ### The protocol layer An LLM on its own does not have access to your live catalog, prices, or inventory. It needs a connection to real data to produce up-to-date and accurate answers. Protocols provide a standardized way for an agent to connect to your commerce backend. Through a protocol, an agent can retrieve real-time data and take actions such as building a cart or placing an order. One emerging standard is the [Universal Commerce Protocol (UCP)](https://developers.google.com/merchant/ucp), an open industry standard. It defines a common language for the shopping journey, including catalog search, cart building, checkout, and order management. AI Hub supports UCP so that agents receive consistent, real-time data. ### The PSP layer The Payment Service Provider (PSP) handles authorization and capture, the same as for any other order. The agent facilitates the transaction, but you remain the merchant of record throughout. Payment details are passed using secure, single-use tokens that are limited to the specific transaction amount. AI Hub integrates with Adyen, Stripe, Worldpay Access, and CyberSource. ### How the layers work together An agentic commerce flow follows these steps: 1. The LLM interprets what the customer wants. 2. The protocol layer retrieves live product, pricing, and inventory data and creates a cart. 3. Shipping and payment details are collected. 4. The PSP authorizes payment. 5. An order is created using your existing pricing, tax, and promotion rules, the same as for any other sales channel. AI Hub connects these three layers to your commercetools Project. Use it to configure which LLM providers, protocols, and PSPs power your agentic commerce channel, so that the components work together without custom integration efforts. ## Supported agentic channels AI Hub capabilities vary by agentic channels, reflecting differences in agentic commerce support and checkout capabilities. Support for agentic channels also depends on the capabilities and requirements of the respective platform providers. These providers may introduce, modify, or discontinue features at their discretion. In addition, merchants may need to meet the provider's eligibility requirements or obtain approval before they can operate on a given agentic channel. ### Google Gemini AI Hub supports the full agentic commerce flow with Google Gemini, from product discovery through checkout, using the Universal Commerce Protocol (UCP). This integration supports the following PSPs for payment: - Adyen - Stripe - CyberSource - Worldpay Access For more information, see the [Google module overview](/agentic-commerce/ai-hub/google/overview.md). ### OpenAI ChatGPT The agentic commerce capabilities of OpenAI ChatGPT are limited to product discovery. This helps customers find and compare products within the conversation without a built-in checkout step. AI Hub supports this discovery integration, so your catalog is surfaced to ChatGPT. End-to-end purchase is not part of that platform's flow. For more information, see the [OpenAI module overview](/agentic-commerce/ai-hub/openai/overview.md). ### Stripe Agentic Commerce Suite For merchants using Stripe as their PSP, AI Hub supports the Stripe Agentic Commerce Suite (ACS). ACS enables agentic checkout across multiple AI agents, including ChatGPT, Gemini, and Perplexity, through a single integration. It connects directly to your commercetools backend and exposes your enriched product data, pricing, and inventory to these agents in a secure, structured way. Orders placed through an agent are processed in Stripe and sent back to commercetools. You manage fulfillment and customer experience the same as for any other order. For more information, see the [Stripe Agentic Commerce overview](/agentic-commerce/ai-hub/stripe/overview.md). ## AI Hub features AI Hub gets your agentic commerce capabilities up and running quickly by providing the following: - **Better product discovery**: GenAI channels are becoming a default product discovery channel for shoppers. For merchants, that means a decrease in website traffic, reduced brand control over their product catalog, and potential revenue decline. AI Hub solves this by enabling brands to send authoritative data into the GenAI channels to capture orders and conversions. - **Simplified setup**: integrating with a new platform generally requires significant engineering resources, new infrastructure, and ongoing maintenance. AI Hub eliminates this complexity by offering easy-to-configure modules and configuration tools. This allows teams to enable agentic commerce with simple setup workflows. Businesses of all sizes can adopt agentic commerce without large upfront infrastructure investments. - **Continuous access to the latest agentic commerce capabilities**: getting started with agentic commerce is only the beginning. AI Hub keeps merchants up-to-date with the rapid changes that LLM providers are expected to make to their chatbots as they continue innovating. commercetools maintains the modules so your technical team doesn't have to. - **Centralized data governance**: by routing all agentic commerce through the AI Hub, merchants maintain a single source of truth for product data, pricing, and business logic. This guarantees consistency across every GenAI channel, no matter which one the customer uses. Set it up once and reuse it across as many channels as you need. - **PSP-agnostic**: AI Hub does not tie you to a single payment provider. You can choose and switch between PSPs based on your needs, regional requirements, or pricing preferences without reworking your core setup. As agentic commerce scales, this lets your checkout experiences adapt to different ecosystems and payment options while your underlying commerce logic stays consistent. ## Benefits of AI Hub Using AI Hub offers several strategic advantages for B2C retailers: - It allows you to reach millions of potential new customers who already use LLMs daily for research, shopping help, and discovery. - By embedding commerce directly into the chat, the friction that often causes drop-offs in the traditional online purchase funnel is reduced. Users can find, compare, and buy products instantly within the same interface, significantly improving conversion rates. - This positions your brand as an innovator in the next era of digital commerce, aligning your business with the growing trend of AI-assisted decision-making. ## Example use cases The following examples show how agentic commerce can work for B2C brands: - **Fashion and apparel**: a customer shopping for apparel can tell an LLM, "I need an outfit for an outdoor wedding reception in San Francisco next month. I like bohemian style, size 8, and my budget is around $350." The AI can instantly synthesize weather data and style trends, curating a complete, perfectly matched look (dress, sandals, jewelry) from the product catalog. The shopper can review the single, optimized cart and complete the order through the chat interface. - **Consumer electronics**: a customer looking for electronics might ask, "Compare the top-rated 65-inch 4K TVs for watching sports." An LLM displays a side-by-side comparison of key specs (refresh rate, brightness, price) and, once a choice is made, allows the customer to add accessories and a warranty. The customer can view shipping estimates and finalize the order without leaving the chat. ## Best practices To get the best results with agentic commerce, follow these best practices: - Complete the [self-paced standalone learning module on preparing your Project for agentic commerce](/learning-prepare-for-agentic-commerce/prepare-for-agentic-commerce/overview.md) to understand how to configure your commercetools Project for success in GenAI channels. - **Maintain high-quality product data**: this impacts your visibility and ranking in LLM product searches. Ensure that your product catalog is well-structured and contains accurate titles, detailed descriptions, and clear images. - **Treat LLMs as an additional commerce channel**: the conversational AI interface adds convenience for your shoppers, but is not a replacement for your own customer support systems. Your brand remains responsible for customer communications, returns, and after-sales services. - **Leverage your existing business logic and validation rules**: these guardrails also apply to transactions initiated through GenAI channels, and exist to maintain consistency and prevent errors. ## Related pages - [Area overview page with navigation](/agentic-commerce.md) - [Next page: Setup and behavior](/agentic-commerce/ai-hub/project-setup-and-api-behavior.md)