Overview

Amazon has rolled out more than 50 upgrades to Rufus, its AI shopping assistant, in what represents the largest enhancement package since the tool's launch. The updates introduce persistent memory across shopping sessions, autonomous purchasing capabilities, visual search, and price-tracking automation. For sellers, these changes significantly alter how customers discover, evaluate, and buy products on the platform.

What's New

• Persistent Memory — Rufus now retains customer preferences and shopping history across sessions, delivering increasingly personalized recommendations over time.
• Autonomous Purchasing — The assistant can independently execute purchases using saved payment methods, handling cart addition and checkout for reorders and routine buys.
• Price Tracking and Auto-Buy — Prime members can set 30-day or 90-day price watches with automated alerts, plus configure target prices that trigger automatic purchases when reached.
• Visual Search — iOS users can upload photos for product identification, handwritten grocery list recognition, and image-based product queries, with Android support expected soon.
• Improved Shopping Intelligence — Underlying models have been upgraded to better understand product categories, evaluate features, and match items to customer intent.
• Contextual Conversations — Rufus maintains full conversational context within sessions and anticipates follow-up questions based on dialogue trajectory.

Persistent Memory Reshapes Product Discovery

Rufus now builds detailed profiles of each customer's household, preferences, and purchasing patterns. If a shopper has previously discussed having young children, owning a pet, or preferring organic products, the assistant stores those details and factors them into future recommendations. Over time, suggestions become more precisely tailored without any additional input from the customer.

For sellers, this shift has real implications. Customers receiving AI-curated recommendations are more likely to see products that genuinely match their needs, which should improve conversion rates and reduce returns for well-matched listings. The flip side is that products lacking accurate attribute data or strong listing optimization may struggle to surface through Rufus-driven recommendations. Comprehensive backend keywords, detailed product attributes, and precise category placement matter more than ever when an AI layer is filtering results on the customer's behalf.