Shopping Companion: A Memory-Augmented LLM Agent for Real-World E-Commerce Tasks

2026-03-16

AI Summary

This work addresses the challenge that existing large language models struggle to capture users’ long-term preferences in e-commerce scenarios and lack a unified end-to-end optimization framework and evaluation benchmark. To this end, we propose Shopping Companion, a lightweight, memory-augmented agent architecture that integrates long-term preference modeling with shopping assistance tasks while supporting user intervention. The framework employs a dual-reward reinforcement learning strategy incorporating tool-level rewards to effectively handle sparse and discontinuous reward signals in multi-turn interactions. Evaluated on the first e-commerce benchmark specifically designed for long-term preference modeling—comprising over one million real product records—our method significantly outperforms strong baselines. Notably, even advanced models such as GPT-5 achieve less than 70% success on this benchmark, underscoring both the task’s difficulty and the efficacy of our approach.

Technology Category

• Machine Learning: Learning Preferences or Rankings
• Humans and AI: Learning Human Values and Preferences
• Multiagent Systems: Mechanism Design

Application Category

• Economics, Online Markets and Human Computation: Humans versus LLMs for data annotation and labeling
• User Modeling, Personalization and Recommendation: Metrics for user behavior and evaluating success
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

In e-commerce, LLM agents show promise for shopping tasks such as recommendations, budgeting, and bundle deals, where accurately capturing user preferences from long-term conversations is critical. However, two challenges hinder realizing this potential: (1) the absence of benchmarks for evaluating long-term preference-aware shopping tasks, and (2) the lack of end-to-end optimization due to existing designs that treat preference identification and shopping assistance as separate components. In this paper, we introduce a novel benchmark with a long-term memory setup, spanning two shopping tasks over 1.2 million real-world products, and propose Shopping Companion, a unified framework that jointly tackles memory retrieval and shopping assistance while supporting user intervention. To train such capabilities, we develop a dual-reward reinforcement learning strategy with tool-wise rewards to handle the sparse and discontinuous rewards inherent in multi-turn interactions. Experimental results demonstrate that even state-of-the-art models (such as GPT-5) achieve success rates under 70% on our benchmark, highlighting the significant challenges in this domain. Notably, our lightweight LLM, trained with Shopping Companion, consistently outperforms strong baselines, achieving better preference capture and task performance, which validates the effectiveness of our unified design.

Problem

• e-commerce
• LLM agent
• long-term preference
• benchmark
• shopping tasks

Innovation

• memory-augmented LLM
• unified framework
• long-term preference
• dual-reward reinforcement learning
• e-commerce benchmark