SAP and Google Cloud unveil agentic commerce architecture

SAP and Google Cloud announced an agentic commerce architecture that connects SAP Commerce Cloud with Google Gemini and Google BigQuery to automate multi-agent marketing and retail operations and to support live inventory checks and personalized engagement across channels.

The design standardizes data exchange using the Universal Commerce Protocol so retailers, payment gateways and autonomous agents can share information and execute retail sequences from search through payment and post-sale resolution without requiring merchants to rebuild backend systems.

A Shopping Assistant built on SAP Commerce Cloud and powered by Google Gemini can be deployed through chat, voice, text interfaces and Google Search AI Mode. The assistant retains state through the shopping cycle and ingests live behavioral inputs, current warehouse capacities and active marketing data so recommendations reflect both customer intent and physical fulfilment capability.

Technical integration uses SAP Engagement Cloud, SAP Business Data Cloud Connect for Google BigQuery and Google Cloud infrastructure. The companies use bidirectional, zero-copy data links with administrative controls so enterprise data stores remain in place rather than duplicated. BigQuery consumes external variables such as weather, location and advertising interaction metrics while SAP Customer Experience systems supply customer profiles, transaction histories, service records and consented engagement data.

The combined data flow enables inventory synchronization: the Shopping Assistant queries live warehouse records before displaying a product, verifies availability and completes cart and payment steps directly against the enterprise backend. The architecture is designed to reduce common failures such as purchases that cannot be fulfilled and checkout interruptions caused by out-of-sync stock data.

Generative AI capabilities from Google Gemini, including a Nano Banana 2 variant, generate localized messaging, images and campaign variations using inputs from BigQuery and SAP systems. Creatives can be delivered through Google Rich Communication Services as interactive messages that update with incoming engagement data. Autonomous agents evaluate campaign performance and adjust content without manual reprogramming of campaign flows.

According to SAP and Google Cloud, the Universal Commerce Protocol and bidirectional links reduce integration costs and speed onboarding into AI-driven channels while keeping retailers’ consented engagement and transaction records in SAP Customer Experience tools. Transaction histories and consented data feed back into local profiles for future interactions handled by Gemini models.

SAP research cited by the companies found 78 percent of firms view AI as essential to customer retention by 2026, while 37 percent share customer data across customer experience platforms and 39 percent share data across CRM platforms.

The companies said they expanded their partnership to address fragmented commerce stacks that rely on inconsistent APIs and to provide a standard protocol for commerce interactions.