Amazon is reportedly developing a new generative AI search engine, codenamed Project Nile, that aims to provide a more personalized, conversational shopping experience for its customers. According to internal documents and sources familiar with the plans, Project Nile will introduce an AI-powered conversational layer on top of Amazon's existing search function to deliver more relevant results, expert recommendations, and detailed product comparisons.
The initiative is a top priority for Amazon's senior leadership, including CEO Andy Jassy, who see enormous potential in using AI to revolutionize the way people shop online. Earlier this year, new Amazon search head Joseph Sirosh wore an "I Love AI" t-shirt to an internal town hall unveiling Project Nile, calling it a confidential effort to bring "the power of large language models" into the shopping experience.
Sirosh envisions the AI assistant being like a knowledgeable in-store salesperson, leveraging data about products and individual customer preferences to provide tailored suggestions and answers to queries. For instance, when a shopper asks for coffee maker recommendations, the AI could provide multiple options for different types of machines, summarize reviews, and make personalized picks based on past purchases and wishlists.
Amazon believes focusing first on mobile could significantly increase sales. While mobile represents nearly 80% of Amazon searches, conversion rates lag far behind the website. An intelligent assistant able to have natural conversations may help users more easily find and evaluate products on phones.
Project Nile reflects intensifying competition in AI search from Microsoft, Google and others. While Amazon has advantages of data and reach, Sirosh acknowledged possible limitations around security and speed to market. Moderation tools and human review processes aim to ensure accuracy and avoid problematic responses.
The launch timeline remains fluid, but Amazon is pushing aggressively on Project Nile, staffing up teams to deliver what Sirosh calls a "future mission we want to aspire to" - an AI that can synthesize Amazon's vast product information to help every customer's individual shopping needs. For the retail giant, success could mean a vital boost of mobile revenue amid slowing growth. But the company will also have to ensure its AI assistant enhances, instead of disrupts, the customer experience.