Search papers, labs, and topics across Lattice.
This paper introduces CatalogAgent, a self-learning system that enhances the accuracy of structured attribute (SA) predictions in e-commerce product catalogs by mediating conflicts between a Generator and Evaluator model. By employing a Supervisor Agent to resolve disagreements and incorporating a Memory Base for continuous learning, the system enables the Generator and Evaluator to improve autonomously without human intervention. The results show significant performance boosts of 15.24% for the Generator and 13.98% for the Evaluator, highlighting a novel approach to context engineering in generative AI models.
A Supervisor Agent can mediate conflicts in LLM outputs, leading to substantial accuracy improvements in e-commerce catalog enrichment.
Product catalogs are the backbone of e-commerce sites, yet a large number of structured attributes (SAs) -- such as material, color, and shape -- often have missing values. Typically, SA values are extracted from product information, including titles and descriptions. While LLM-based generator-evaluator frameworks have demonstrated effectiveness for SA prediction -- where an LLM generates SA values and another evaluates them -- they face challenges when the Generator and Evaluator produce conflicting outputs, as either component can make mistakes. We introduce \texttt{CatalogAgent}, a novel agentic system that continuously improves Generator and Evaluator models for e-commerce catalog enrichment. When disagreements arise from (1) internal conflicts between the LLM-based Generator and Evaluator, or (2) external feedback from sellers on LLM outputs, a Supervisor Agent intervenes to mediate these conflicts and make final decisions. The system also incorporates a Memory Base and a Memory Summarizer that stores Supervisor Agent activities from individual cases and aggregates patterns into learnings. These learnings are fed back to the worker Generator and Evaluator LLMs, enabling self-improvement without human intervention. Through context engineering -- injecting learnings and insights into worker LLMs'contexts -- the system successfully transfers the Supervisor's capabilities to the Generator and Evaluator, improving their performance by 15.24\% and 13.98\%, respectively. Our experiments demonstrate a new paradigm of Supervisor Agent-mediated self-learning systems for improving generative AI model accuracy.
