Amazon Rufus Is Now Alexa for Shopping: Build Your Own AI Shopping Assistant
Amazon’s Alexa for Shopping, formerly Rufus, shows how product search, personal advice, and purchase flows can be combined in a single conversation. With the LoyJoy Platform, companies apply this same principle to their own store, their own brand, and their own customer data.
An AI Shopping Assistant does not just answer product questions. It understands the occasion, asks targeted follow-up questions, recommends suitable products, and guides the customer to the next step in the purchase process.
With LoyJoy, business teams build AI Shopping Assistants themselves, with no coding required. Product data, the shop system, CRM, and cart can be connected via APIs and existing integrations.
See an AI Shopping Assistant in a demo · Try LoyJoy free for 30 days
Rufus is now Alexa for Shopping
Amazon transitioned Rufus into Alexa for Shopping on May 13, 2026. This is more than a rename. Amazon combines Rufus’s product knowledge and purchase history with the personal, cross-device context of Alexa+.
Alexa for Shopping is thus a personal shopping assistant that supports customers through research, selection, and ordering. It is currently focused primarily on the Amazon ecosystem and the US market.
For other companies, Amazon’s technical scale matters less than the underlying pattern:
- Customers state their goal in natural language.
- The assistant asks follow-up questions where needed.
- Product, customer, and context data are brought together.
- The assistant recommends and justifies suitable options.
- Shop systems reliably execute cart, order, or notification actions.
What Alexa for Shopping can really do
1. Find products by goal, not just keyword
Customers do not need to know a product category or an exact search term. They can describe, for example, that they are looking for a skincare routine, gear for a trip, or a gift for a specific person.
2. Explain categories and products
Alexa for Shopping answers questions about products, features, and use. For larger purchases, the assistant creates individual buying guides and explains which features matter for the decision.
3. Compare products
Selected products can be compared by features, price, and reviews. The customer gets more than a results list; they get support for the actual decision.
4. Personalize recommendations
Amazon takes past purchases, search behavior, and explicitly stated preferences into account, for example interests, dietary needs, family members, or pets. Customers can review and correct what the assistant has stored about them.
5. Understand images, text, and voice
Customers can write, speak, or upload images freely. Rufus, for example, could recognize a handwritten shopping list or search for similar items based on a product photo.
6. Incorporate customer data and orders
The assistant answers questions about order status, looks up past purchases, reassembles frequently bought items, and adds products to the cart.
7. Track prices and prepare actions
Alexa for Shopping shows price history, sets price alerts, and can, under certain conditions, execute a purchase at a defined target price. Recurring purchases and notifications can be set up as scheduled actions.
8. Provide customer service
The assistant also helps with delivery status, returns, refunds, damaged items, changes, and account questions. It hands over to customer service when needed.
What does your own online store actually need?
Not every store needs the full scope of Amazon’s feature set right away. For most companies, the greatest business value comes from five capabilities:
- Reliably answering product questions
- Understanding and qualifying needs
- Recommending suitable products or routines
- Taking current prices and availability into account
- Guiding customers directly into cart, checkout, or advisory flows
Price history, automatic purchases, and long-term customer memory are possible later stages. They require additional matching data, consent, and reliable backend processes.
How LoyJoy implements an AI Shopping Assistant
| Function | Implementation with LoyJoy |
|---|---|
| Product and usage questions | AI Agent with RAG over websites, product information, guides, and documents |
| Individual advice | Free-form dialogue, targeted follow-up questions, questionnaires, and decision logic |
| Product recommendations | AI Product Recommender with product search and product gallery |
| Current product data | Product database, CSV, PIM, shop API, database, or JSON interface |
| Product comparison | Comparison flow, product gallery, or a custom web component |
| Prices and availability | Live lookup from the shop, PIM, or ERP via API |
| Cart | Add-to-cart link or direct connection to the shop system |
| Customer account | Sign-in via one-time code or integration with Auth0, Keycloak, and existing identity systems |
| Order history and preferences | Lookup and storage via CRM, CDP, or customer account |
| Image-based advice | Image upload plus a connected vision model or the customer’s own review service |
| Advice by voice | LoyJoy Phone Agent or voice input in the digital channel |
| Notifications | Consent-based email, web push, or the customer’s own notification processes |
| Human advice | Handover to live chat, customer service, or a specialist with context |
LoyJoy does not replace the shop, PIM, or ERP system. The AI Agent understands the request, drives the dialogue, and calls the appropriate functions. Binding orders, payments, and inventory changes continue to be executed by the systems designed for that purpose.
AI Agent plus reliable business processes
Amazon, too, separates conversation from execution. For a price alert, the AI Agent creates a structured task. A classic software service then monitors the price and executes the defined action.
The same architectural principle applies to LoyJoy:
- The AI Agent understands the goal, conditions, and customer input.
- LoyJoy captures the required data and consent.
- APIs hand a structured task to the shop, CRM, or backend.
- The connected system executes the action in a controlled way.
- LoyJoy communicates the result, status, or next step.
This is how LoyJoy connects natural conversation with traceable, controllable business processes.
La Biosthétique: Personal product advice in the online store
La Biosthétique brings the personal advisory quality of trained partner salons into its own online store. The AI product advisor on the LoyJoy Platform is available directly on product pages, exactly where purchase decisions are made.
The assistant offers customers several entry points:
- Self-assessment and individual diagnosis
- Questions about products and use
- Building a personal care or styling routine
- Newsletter sign-up
- Finding a partner salon for professional in-person advice
Styling, skincare, and coloration each follow different advisory logic. That is why La Biosthétique uses specialized advisory templates instead of a single generic flow.
From the answers, the assistant assembles an individual routine and shows the recommended products with image, description, and price directly in the dialogue. Complementary products make complete routines visible and support cross-selling. Higher-value alternatives enable targeted up-selling.
Read the La Biosthétique case study · See the AI product advisor live
Four stages to your own AI Shopping Assistant
Stage 1: Product knowledge. The assistant answers questions about products, features, use, and compatibility.
Stage 2: Individual product recommendation. The assistant understands occasion, preferences, and budget, asks follow-up questions, and recommends suitable products.
Stage 3: Shop and customer integration. Current prices, stock, customer account, order history, and cart are connected.
Stage 4: Proactive and recurring actions. With customer consent, reminders, reorders, price notifications, or personalized offers are implemented.
A first prototype can be built in LoyJoy within minutes. The scope of the production rollout depends on the extent and availability of the systems to be connected.
What LoyJoy deliberately does not promise
LoyJoy does not automatically replicate the entire Amazon ecosystem. Features such as a cross-device Alexa memory, a full year of price history, or automatic purchases with stored payment methods require corresponding customer data and backend services.
The advantage for companies lies elsewhere: they get their own, brand-consistent assistant with control over knowledge, recommendations, data, processes, and integrations.
Frequently asked questions about the AI Shopping Assistant
What is an AI Shopping Assistant?
An AI Shopping Assistant advises customers in natural language during product search and purchase decisions. It answers product questions, asks follow-up questions, recommends suitable items, and can call shop functions such as the cart or order status.
Is Amazon Rufus now Alexa for Shopping?
Yes. Amazon transitioned Rufus into Alexa for Shopping on May 13, 2026. The new offering combines Rufus’s shopping capabilities with the personal, cross-device context of Alexa+.
Can I integrate Alexa for Shopping into my own store?
Alexa for Shopping is an Amazon product. With LoyJoy, however, companies can build their own AI Shopping Assistant based on a comparable functional principle for their own brand, products, and systems.
Which shop systems can be connected to LoyJoy?
LoyJoy connects to shop, PIM, CRM, and ERP systems via existing integrations, add-to-cart links, and REST APIs, including Shopify, Shopware, Magento, and WooCommerce.
How does the assistant get current product data?
Product data can be provided via CSV, API, database, JSON, or PIM. Prices and availability can be queried live from the shop or ERP system if needed.
Can the assistant add products directly to the cart?
Yes, via an add-to-cart link or a direct shop API. The actual checkout and payment remain within the shop system.
Can LoyJoy take past purchases and customer preferences into account?
Yes, once customers are identified and the relevant data is provided from the customer account, CRM, or CDP. Storage and use must comply with the company’s consent and data protection concept.
Does the AI Shopping Assistant also work by voice?
Yes. LoyJoy supports AI Agents by chat and phone, so the same business logic can be used across digital channels and service or advisory telephony.
Start building your own AI Shopping Assistant
Build a first prototype with LoyJoy and test the advisory flow with your own products. For production use, we connect product data, the shop system, and customer processes into one seamless experience.
Try LoyJoy free for 30 days · Request a personal demo
For a related read, see how companies build their own AI Customer Service Agent inspired by CHECK24 Sophie.
Note: Amazon, Alexa, and Rufus are trademarks of their respective owners. LoyJoy is not affiliated with Amazon. This article is provided for factual comparison of different functional principles.
Sources: