Amazon’s new AI shopping tool tells you why you should buy a recommended product (techcrunch.com)

Amazon introduced “Help me decide,” an AI-driven shopping assistant that analyzes your Amazon searches, browsing and purchase history to recommend products and explain why each pick fits your needs. The feature appears after you’ve viewed multiple similar listings or under “Keep shopping for” on the homepage and initially respects the price range you’re browsing (with options to expand to cheaper or pricier items). For example, if you’ve been looking at four-person sleeping bags, stoves and bought camping boots, it might recommend a four-person all-season tent and describe why that choice matches your past activity. Technically, Amazon combines large language models with AWS services — Bedrock for generative AI, OpenSearch for search, and SageMaker for recommendations — to fuse retrieval, personalization and natural-language justification. For the AI/ML community this is a clear example of productionizing LLMs alongside traditional search-and-recommendation systems to drive conversions, showing how RAG-like architectures and model-service integration can surface personalized, human-readable rationales. The rollout on iOS, Android and web underscores the push to embed generative explanations into UX, raising opportunities for richer evaluation metrics (explainability, trust, lift) and questions about data privacy and alignment between user intent and model suggestions.