Your SAP Commerce Cloud Just Got an AI Brain — The Complete Guide to SAP CX AI Toolkit

 

6 AI-Powered Tools Built into Commerce Cloud

GA Shopping Agent Available Since H1 2025

50+ Languages Supported For Content Generation

100% Native SAP Integration No Third-Party Middleware

The Gap Between Browsing and Buying Is Costing You More Than You Think

Every e-commerce leader knows the numbers. The average conversion rate across B2C and B2B commerce sits at 2–4%. The average cart abandonment rate hovers above 70%. Customers who cannot find the right product in under three clicks leave. Product descriptions that are incomplete, poorly written, or missing in non-English languages are invisible to both search engines and shoppers. Customer service teams spend the majority of their day answering questions a machine could resolve in seconds.

These are not technology problems. They are AI-readiness problems. And for SAP Commerce Cloud customers, the solution is already embedded in your platform: the SAP CX AI Toolkit.

Shopping is no longer about transactions. It's about intelligent, connected experiences that understand what customers need before they even say it.

Released broadly in 2024 and achieving general availability in H1 2025, the SAP CX AI Toolkit is not a bolt-on AI chatbot or a third-party integration. It is a native, governed collection of AI agents and tools built directly into SAP Commerce Cloud — designed to transform every layer of the commerce experience, from product data to the purchase decision.

This blog breaks down every capability, what it does in practice, and what it means for your revenue, your team, and your competitive position.

What Is the SAP CX AI Toolkit — And Why Is It Different?

Let's be direct: the enterprise software market is saturated with AI add-ons, point solutions, and experimental pilots that require expensive custom integration, deliver inconsistent results, and quietly disappear from roadmaps when the hype cycle moves on.

The SAP CX AI Toolkit is none of those things. It is SAP's native AI layer for its entire Customer Experience portfolio — built on SAP Business Technology Platform (BTP), powered by SAP Business AI, and governed by the same security, compliance, and data architecture that SAP's enterprise customers depend on for their most critical operations.

What Makes It Architecturally Different

The SAP CX AI Toolkit connects directly to your SAP Commerce Cloud product catalog, your customer data, your order history, and your content structures. AI agents do not work on a copy of your data or a disconnected API feed. They operate on the same live data that powers your storefront — meaning product descriptions are generated from the actual attributes in your catalog, the Shopping Agent draws on real-time inventory, and the Catalog Optimization Agent assesses the real state of your content quality. This is governed, grounded AI — not generative guesswork.

For SAP Commerce Cloud customers, the toolkit is accessed through SAP CX AI Toolkit, a centralised hub where administrators configure AI agents, set access controls, monitor pipelines, and deploy capabilities to the storefront and back office. The toolkit connects natively to SAP Sales Cloud, SAP Service Cloud, SAP Emarsys, and the SAP Customer Data Platform — creating a unified AI layer across the entire customer experience.

The Six AI Powers Inside SAP Commerce Cloud

1) The AI Shopping Agent — Your Always-On Digital Sales Expert

The Shopping Agent is the flagship AI experience of the SAP CX AI Toolkit. It transforms your storefront from a product catalogue into a conversational selling environment. Instead of forcing customers to navigate menus, filter results, and decode technical specifications — the Shopping Agent simply asks what they need, and answers.

AI COMMERCE

Shopping Agent — Conversational Commerce at Scale

Generally available since H1 2025 | Natural language | Multi-language | Brand-tone configurable

• Natural language product search — customers describe what they need, the agent returns exact matches.
• Complex specification queries — answers questions hidden in product manuals, PDFs, and data sheets.
• Personalised recommendations — surfaces relevant alternatives and upsells based on the conversation.
• Add-to-Cart from conversation — seamlessly moves from discovery to purchase without leaving the chat.
• Stock-aware recommendations — hides out-of-stock items in real time; supports B2B bulk availability queries.
• Order status subagent — customers can check their orders without leaving the shopping conversation.
• Brand-tone configuration — choose the agent's tone to match your brand voice (formal, consultative, friendly).
• Multi-language — supports all languages available in the Commerce Backoffice, with auto-detection.
• Combinable with Digital Service Agent — one unified conversational surface for shopping and support.

 

2) Product Description Generator — Catalog Content at the Speed of AI

Your product catalog is your most important sales asset — and for most commerce teams, it is also their biggest bottleneck. Writing, translating, and maintaining high-quality descriptions for thousands of SKUs is a task that is never truly finished. The Product Description Generator changes the equation entirely.

PD CONTENT

Product Description Generator — Bulk, Branded, Multilingual

Available now | Templates for brand consistency | Bulk generation | 50+ languages | Direct publish to catalog

• Generates descriptions using live product attributes directly from your SAP Commerce Cloud catalog.
• Custom prompt templates — define your brand voice, style, length, and tone guidelines once, apply everywhere.
• Bulk generation — process thousands of SKUs simultaneously; prioritise gaps or rewrites.
• Classification attribute value generation — automatically populates missing attributes from product context.
• Multi-language output — generate content in all Commerce Backoffice languages; no manual translation workflow.
• Direct publish — AI-generated content moves straight to your catalog with one click, no copy-paste.
• SEO-optimised — descriptions structured for search discoverability, not just human readability.
• Translation AI Agent — separate tool for translating existing product names, descriptions, and attributes at scale.

3) Catalog Optimization Agent — Your AI Merchandising Manager

The Catalog Optimization Agent (in Beta since November 2025) is the closest thing to an autonomous merchandising manager that exists in the SAP ecosystem. While your team is focused on strategy, campaigns, and commercial decisions, this agent is silently auditing your entire product catalog against predefined quality standards — and surfacing exactly what needs to change and why.

CO CATALOG

Catalog Optimization Agent — Autonomous Catalog Quality Management

Beta since November 2025 | Analyzes descriptions, attributes, translations | Generates recommendations

• Audits product descriptions, attributes, and translations against your quality standards.
• Identifies gaps — missing attributes, incomplete descriptions, below-standard translations.
• Generates actionable recommendations — not just flags, but suggested improvements ready to review.
• AI Insight Cards — generates commerce intelligence cards from catalog and performance data (Q3 2025+).
• Configurable thresholds — administrators set the quality bar and the scope of the audit.
• Directly connected to the Product Description Generator — recommendations feed directly into content creation.

4) Visual Search — Find Products With a Photo

Visual search is one of the most commercially powerful features in the entire toolkit — and one of the most underestimated. For retail, fashion, furniture, electronics, and any product category where aesthetics matter, it closes the gap between a customer seeing something they want and finding it in your catalog.

VS DISCOVERY

Visual Search — Image-to-Product Discovery

Available now | Upload any image | Visual Information Extraction | Fashion, furniture, electronics

• Image-based product search — customer uploads a photo of any product or scene and gets matched results.
• Works with specific products (a camera, a pair of shoes) or general scenes (a dressed model, a furnished room).
• Finds visually similar products — matches colour, shape, style, and category from the uploaded image.
• Visual Information Extraction — extract structured data from product images using natural language instructions.
• Enhanced product Q&A — Ask About This Product, now enriched with answers from product reviews.
• Review Summariser — generates AI summary of all reviews for any product, surfaced directly on the PDP.

5) Product Image AI — Generate, Edit, and Publish Visuals at Scale

For commerce teams managing thousands of SKUs, creating and maintaining high-quality product imagery is a constant operational burden. Photoshoots are expensive. Re-shoots for minor changes are impractical. Localising images — adding regional labels, removing background objects, updating overlays — consumes designer hours that should be spent on strategic creative.

IM IMAGES

Product Image AI — Generate, Edit and Localise Product Visuals

Available now | Generate new images | Edit existing images | Add overlays | Direct publish to catalog

• AI image generation — generate product images from text prompts, in any supported language.
• Object removal and replacement — remove or replace specific objects in existing product images with AI.
• Overlay addition — add logos, labels, price tags, or promotional text as overlays to existing images.
• Visual Information Extraction — extract structured product data from image content using instructions.
• Direct publish — AI-generated images publish straight to the Commerce Cloud catalog; no manual upload.
• Multi-language prompt support — generate and edit images using prompts in any supported language.

6) Digital Service Agent + Joule Studio Agent Builder — The Automation Layer

The SAP CX AI Toolkit is not limited to storefront customer experiences. It extends into service operations and custom business logic — giving commerce and operations teams a powerful automation layer that reduces support costs, accelerates resolution, and connects AI to the broader business.

SA AGENTS

Digital Service Agent + Joule Studio Agent Builder

Available now | Deflect common inquiries | Low-code/no-code custom agents | Email automation | API call triggers

• Digital Service Agent — AI-powered agent that deflects common customer inquiries based on your knowledge base.
• Seamless handoff — when escalation is needed, the agent transfers the case with full context to a human agent.
• Combined with Shopping Agent — one unified conversational surface for shopping and service queries.
• Joule Studio Agent Builder — part of SAP Build, a low-code/no-code environment to create custom AI agents tailored to your business processes (agent builder GA in Q1 2026).
• Automated email actions — custom agents can trigger automated email sends based on business rules.
• API call actions — custom agents can create records in SAP and external applications via API.
• Data sources — pull from SAP Sales Cloud, Service Cloud, Commerce Cloud, and Customer Data Platform.
• 360-degree customer view — custom tools can access full customer profile for personalised automation.
• Role-based access controls — administrators manage who builds, tests, and deploys custom agents.

 

The Commerce Experience, Transformed — Before and After

One Toolkit. Your Entire Customer Experience Ecosystem.

What makes the SAP CX AI Toolkit genuinely powerful is not any individual capability — it is the integration architecture that connects those capabilities across your entire technology landscape. This is not a siloed e-commerce AI feature. It is the AI layer for SAP's full Customer Experience suite.

Commerce + ERP: The Data Advantage

Because the SAP CX AI Toolkit is native to SAP's architecture, the Shopping Agent does not just draw on product catalog data. It can surface real-time inventory, pricing, order status, and customer-specific contract terms from your SAP S/4HANA or ERP backend — giving B2B buyers the kind of accurate, contextual responses that no third-party AI tool could provide without expensive custom integration.

Commerce + CDP: Hyper-Personalised Automation

Custom AI agents built in the Agent Builder can pull from the SAP Customer Data Platform — accessing full customer profiles, purchase history, segment data, and behavioural signals. This enables personalised automation flows that adapt to individual customers, not just product segments.

Commerce + Marketing: AI-Assisted Campaigns

The toolkit connects to SAP Emarsys for AI-assisted campaign creation, including AI-powered product finder for campaign targeting, automated campaign translation, and AI-generated promotion recommendations — bridging the gap between catalog intelligence and marketing execution.

External Systems via API

Custom agents built in the no-code Agent Builder can make API calls to external systems — creating records, triggering workflows, and exchanging data with Salesforce, Microsoft Dynamics, third-party PIM systems, and any system with a REST API. SAP CX AI Toolkit also supports Azure Blob Storage integration for external data pipelines.

What Is Coming: The SAP CX AI Toolkit Roadmap

  Figure 4: SAP CX AI Toolkit for Commerce — Feature Rollout Timeline

 

SAP's commitment to the CX AI Toolkit is clear from the release cadence: significant new capabilities ship every quarter, with the roadmap explicitly tied to SAP's broader Autonomous Enterprise and Joule strategy. Here are the most commercially important developments ahead:

• Full Joule Integration (H2 2026+): The Joule Studio agent builder reached general availability in Q1 2026, and the Shopping Agent and Catalog Optimization Agent are progressively surfaced within Joule — making CX agents accessible through SAP's unified agentic workspace alongside finance, HR, and supply chain agents.
• Expanded AI Insight Cards: AI-generated commerce intelligence cards based on catalog performance, customer interaction data, and agent conversation history — giving merchandisers and category managers actionable insight without building custom dashboards.
• AI-Assisted Promotion Creation: Recommendations for promotion names, timing, product selection, discount type and spend category — reducing the time from promotion idea to live campaign.
• Extended Custom Agent Capabilities: Custom agents that can send automated emails, create records in SAP and external systems, and orchestrate multi-step workflows across the CX portfolio.
• Granular Role-Based Access Controls: User-level and role-level controls for all AI tools and pipelines — addressing enterprise governance and compliance requirements as agent usage scales.
• Multi-CX Data Combination: Agents that can draw simultaneously from Commerce, Sales, Service, and CDP data — enabling truly unified customer intelligence across the entire buying and service journey.

The Business Case: What AI-Powered Commerce Actually Delivers

Marketing language aside — let's talk about what activating the SAP CX AI Toolkit actually moves on your P&L. These are the commercial levers that the toolkit directly influences:

Conversion Rate

The Shopping Agent reduces search abandonment — one of the single biggest drivers of lost conversion. By intercepting customers who cannot find what they need with a conversational guide, it brings them to the right product. Add-to-Cart from conversation eliminates the friction of navigating from discovery to purchase. Visual search captures intent that text search misses entirely. The combined impact across these three mechanisms is measurable and fast — SAP's Q1 2025 release cites an approximately 5% increase across online conversion rate, average order value, and repeat purchases from the AI shopping assistant.

Average Order Value

An AI Shopping Agent that understands context can recommend complementary products, upgraded specifications, or bundle options during the conversation — at exactly the moment of maximum purchase intent. This is contextual upsell that operates at a scale no sales team could match.

Catalog Operational Costs

Product description writing, translation, attribute population, and image creation are among the highest-volume, lowest-value activities in a commerce operations team. The Product Description Generator, Translation Agent, and Product Image AI collectively eliminate or dramatically reduce the manual effort in each of these areas — freeing team capacity for strategic merchandising, campaign planning, and commercial analysis.

Customer Service Cost

The Digital Service Agent deflects common product and order queries — the high-volume, low-complexity requests that consume the most agent time. When combined with the Shopping Agent and integrated with the SAP Service Cloud handoff capability, it creates a seamless self-service experience that reduces inbound contact volume without sacrificing customer satisfaction.

SEO and Organic Traffic

Product pages with rich, complete, accurate, and keyword-relevant descriptions rank better. Pages with multiple languages serve more markets. Pages that answer questions customers actually ask (enriched from PDFs and reviews) increase dwell time and reduce bounce rate. The Product Description Generator and catalog enrichment tools are, in effect, a scalable SEO programme.

This toolkit is not a cost centre — it is a revenue driver. Every capability maps directly to a commercial outcome: higher conversion, lower operational cost, broader market reach, or reduced churn.

How Akkodis Activates the SAP CX AI Toolkit — End to End

The SAP CX AI Toolkit is powerful out of the box — but 'out of the box' and 'delivering commercial value' are not the same thing. The organisations that extract the most from the toolkit are those with a structured activation approach: the right features for the right use cases, configured to their catalog, trained on their brand voice, and governed for their compliance requirements.

That is exactly what Akkodis SAP Practice delivers. As a recognised SAP CX partner, we have deep expertise across the full CX AI Toolkit — and we operate from a commercial mindset, not an IT one. Our goal is not to install features. It is to deliver outcomes.

Phase 1 — AI Readiness & Use Case Prioritisation (2–3 Weeks)

We assess your current SAP Commerce Cloud environment — catalog data quality, content gaps, language coverage, customer service volumes, and storefront performance — against the capabilities of the CX AI Toolkit. The output is a prioritised activation plan: the specific tools that will deliver the fastest, largest, most measurable commercial impact for your business.

Phase 2 — Toolkit Configuration & Catalog Preparation (3–6 Weeks)

Configuration of the CX AI Toolkit against your Commerce Cloud environment — connecting AI agents to your catalog, setting brand-tone guidelines, configuring language rules, establishing access controls, and preparing the product data quality baseline that AI agents need to produce accurate results.

Phase 3 — Prompt Engineering & Brand Voice Training (2–4 Weeks)

Generic AI produces generic results. We build the custom prompt templates, tone configuration, and content guidelines that make the Product Description Generator and Shopping Agent sound like your brand — not like every other commerce platform. This phase includes testing across representative SKU samples and storefront UX validation.

Phase 4 — Deployment, Testing & Go-Live (2–4 Weeks)

Phased deployment with A/B testing where appropriate — allowing you to measure the direct impact of each AI capability against your baseline before full rollout. We instrument the key commercial KPIs from day one: conversion rate, cart abandonment, support deflection, content production speed.

Phase 5 — Optimisation, Expansion & Ongoing Support

Post-launch optimisation — refining prompts, expanding language coverage, onboarding additional AI tools as they reach GA, and extending the toolkit into adjacent use cases (sales, service, marketing) as your confidence and capability grows.

Typical Akkodis CX AI Toolkit Activation Timeline

Initial Shopping Agent + Product Description Generator activation: 6–10 weeks from kick-off to go-live. Full toolkit activation including Catalog Optimization Agent, Visual Search, and Custom Agents: 12–18 weeks. All timelines include configuration, testing, brand-voice training, and commercial KPI instrumentation.

Intelligence Is the New Competitive Advantage in Commerce.

Your competitors are already exploring SAP CX AI Toolkit. The brands that activate it first — and activate it right — will own the customer experience for the next decade. Akkodis SAP Practice can take you from evaluation to live deployment in weeks, not months.