Cleo: A Transparent and Controllable Chatbot for Conversational Commerce

2026-08-06

AI Summary

This work addresses the opacity, unpredictability, and difficulty of large language models (LLMs) in handling complex product comparisons within conversational e-commerce by proposing a transparent and controllable dialogue recommendation system. The system integrates a deterministic ranker—based on categorical filtering over 3,638-dimensional product specifications and an interpretable numerical loss function—with a constrained LLM that generates descriptions strictly from the product catalog. This architecture enables requirement clarification, real-time re-ranking, and natural language comparisons across multiple items. By decoupling ranking from text generation, visualizing attribute-wise loss explanations, and incorporating a highlighting mechanism for comparative features, the system enhances algorithmic transparency while preserving conversational fluency. Consequently, it significantly reduces users’ cognitive load during decision-making and supports auditable, scalable, and explainable recommendations.

Abstract

We demonstrate Cleo, a transparent and controllable conversational product advisor that addresses the challenges of opacity, unpredictability of LLMs, and the complexity of comparisons in conversational commerce. With our chatbot system, we make four contributions: First, we introduce transparency by prompting the LLM to reflect on interpreted user needs, while an auditable ranking mechanism reveals loss values per attribute, explaining ranking decisions. Second, we propose controllability through a hybrid architecture separating deterministic ranking from language generation. A ranker applies categorical filters and numeric loss functions over 3,638 product specifications. Meanwhile, a constrained LLM generates grounded descriptions constrained to catalog evidence, thus mitigating the risk of hallucinated or persuasive content. Third, we provide decision support in the form of natural-language comparisons and a highlights feature. These aim to reduce mental workload by contextualizing specifications relative to user needs. Fourth, we contribute an extensible experimental system for IR and HCI researchers, as well as practitioners of conversational search and recommendation. Unlike traditional faceted search or opaque LLM-only recommenders, our approach allows for fluid conversation while maintaining algorithmic transparency. In a live demonstration, attendees will experience information needs elicitation and reflection, conversational refinement with real-time re-ranking, inspection of per-attribute loss explanations, and AI-generated multi-item comparisons. The system aims to advance the design of transparent and controllable conversational systems that provide support for decision-making during online product search.

Problem

• conversational commerce
• opacity
• unpredictability
• comparison complexity
• decision support

Innovation

• transparency
• controllability
• conversational commerce
• hybrid architecture
• explainable ranking