Amazon Introduces Rufus, a New AI-Powered Shopping Assistant in Its Mobile App

Home AI Projects Amazon Introduces Rufus, a New AI-Powered Shopping Assistant in Its Mobile App
Amazon Introduces Rufus, a New AI-Powered Shopping Assistant in Its Mobile App

Amazon today unveiled an innovative virtual shopping companion called Rufus, equipped with artificial intelligence technology to navigate the retailer’s wide range of products and integrate information gathered from the Internet. The digital assistant is set to debut in beta for a limited number of users in the US starting today, and is expected to roll out to a wider audience in the coming weeks. Rufus will be integrated into Amazon’s mobile app, allowing customers to ask for help finding products, compare and receive personalized recommendations.

 

This cutting-edge AI chatbot is the latest in a series of technological improvements that Amazon has introduced to enhance the customer experience. These enhancements include a variety of tools designed to assist with outfit selection, aggregate product reviews, and even support advertisers and sellers.

 

Built on a comprehensive learning regimen using Amazon’s vast product portfolio, customer reviews, community engagement, and additional web resources, Rufus is well-equipped to handle a wide range of shopping inquiries. Whether shoppers are starting to shop, narrowing down their choices, or looking for answers to specific questions about a product, Rufus is ready to provide guidance.

 

According to TechCrunch, Amazon has developed an exclusive large-scale language model (LLM) of its own about the shopping experience, training it on both Amazon data and a set of public Internet data. The exact scope of the external data sources has not been disclosed, including whether they include other retail websites. For example, Amazon suggests that a shopper interested in purchasing running shoes might ask Rufus about important factors to consider, the differences between trail and road running shoes, or the durability of a particular pair. Inquiries for other product categories might include what to look for in a set of headphones, tips for taking care of your car at home, identifying clean beauty products, cold-weather golfing essentials, or even creating an indoor garden.

 

Rufus also shines when tasked with comparing products or when customers ask for advice, such as Valentine’s Day gift suggestions or choosing the best dinosaur-themed toys for a child. By following Rufus’ answers, customers can dig deeper into more curated search results. Similar to interactions with other consumer AI chatbots, such as OpenAI’s ChatGPT or Google’s Bard, which also integrates shopping features, users are free to interact with Rufus.

 

Rufus will initially debut in beta for select US customers via a new button in the bottom navigation bar of the Amazon app. Buyers have the option to type or speak their requests, which activates a chat window at the bottom of the screen. Once the interaction is complete, customers can simply swipe down on the screen to collapse the chat window and return to their normal browsing of Amazon apps.

 

Amazon expects the beta phase to contribute to ongoing improvements to Rufus and its broader generative AI projects. “In the nascent stage of generative artificial intelligence, the technology may not always be flawless,” Amazon said in a recent blog post. “Our commitment is to constantly improve our AI models and adjust the answers, gradually increasing the usefulness of Rufus. Customer feedback is invaluable, and users can express their satisfaction or dissatisfaction through the rating system, with the option to leave additional comments,” it continues.

 

Amazon confirms that Rufus will debut ad-free, but additional features may be included in the experience to further meet customer needs. The effectiveness of Rufus remains untested as it was not available for review. However, it’s important to note that Amazon’s previous Q chatbot, aimed at businesses, faced issues related to misinformation and disclosure of sensitive data.

allix