WISE-Flow: Workflow-Induced Structured Experience for Self-Evolving Conversational Service Agents

2026-01-13

AI Summary

This work addresses the instability of large language model (LLM) agents in novel tasks, which often stems from repetitive errors and inconsistent internal states. Conventional repair strategies are typically costly and lack scalability. To overcome these limitations, the authors propose WISE-Flow, a novel framework that introduces a workflow-centric approach to structured experience modeling. WISE-Flow extracts action blocks with explicit preconditions from historical interactions to construct reusable workflow experiences. During inference, it aligns execution trajectories and performs feasibility reasoning to generate state-consistent next actions. Experiments demonstrate that this method significantly improves both task success rates and execution stability across multiple base models on ToolSandbox and τ²-bench benchmarks.

Technology Category

• Multiagent Systems: Teamwork
• Humans and AI: Human-Aware Planning and Behavior Prediction
• Planning, Routing, and Scheduling: Planning with Language Models

Application Category

• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• Economics, Online Markets and Human Computation: Architectures and workflows that use LLMs for crowd work
• User Modeling, Personalization and Recommendation: Large Language Models (LLM) for user modeling and recommendation

Abstract

Large language model (LLM)-based agents are widely deployed in user-facing services but remain error-prone in new tasks, tend to repeat the same failure patterns, and show substantial run-to-run variability. Fixing failures via environment-specific training or manual patching is costly and hard to scale. To enable self-evolving agents in user-facing service environments, we propose WISE-Flow, a workflow-centric framework that converts historical service interactions into reusable procedural experience by inducing workflows with prerequisite-augmented action blocks. At deployment, WISE-Flow aligns the agent's execution trajectory to retrieved workflows and performs prerequisite-aware feasibility reasoning to achieve state-grounded next actions. Experiments on ToolSandbox and $\tau^2$-bench show consistent improvement across base models.

Problem

• LLM-based agents
• error-prone
• failure patterns
• run-to-run variability
• self-evolving

Innovation

• workflow-induced experience
• self-evolving agents
• prerequisite-augmented action blocks
• feasibility reasoning
• structured experience