MindFlow+: A Self-Evolving Agent for E-Commerce Customer Service

TL;DR

MindFlow+ is a self-evolving dialogue agent for e-commerce customer service that combines large language models with imitation and reinforcement learning to improve response quality and task accuracy in dynamic multi-turn interactions.

Contribution

The paper introduces MindFlow+, a novel self-evolving agent that integrates tool-augmented demonstration and reward-conditioned data modeling for domain-specific dialogue learning.

Findings

Outperforms baselines in contextual relevance and flexibility

Effectively uses tool reasoning and reward signals for better responses

Demonstrates improved task accuracy in real-world conversations

Abstract

High-quality dialogue is crucial for e-commerce customer service, yet traditional intent-based systems struggle with dynamic, multi-turn interactions. We present MindFlow+, a self-evolving dialogue agent that learns domain-specific behavior by combining large language models (LLMs) with imitation learning and offline reinforcement learning (RL). MindFlow+ introduces two data-centric mechanisms to guide learning: tool-augmented demonstration construction, which exposes the model to knowledge-enhanced and agentic (ReAct-style) interactions for effective tool use; and reward-conditioned data modeling, which aligns responses with task-specific goals using reward signals. To evaluate the model's role in response generation, we introduce the AI Contribution Ratio, a novel metric quantifying AI involvement in dialogue. Experiments on real-world e-commerce conversations show that MindFlow+ outperforms strong baselines in contextual relevance, flexibility, and task accuracy. These results demonstrate the potential of combining LLMs tool reasoning, and reward-guided learning to build domain-specialized, context-aware dialogue systems.