WISE-Flow: Workflow-Induced Structured Experience for Self-Evolving Conversational Service Agents
This addresses the costly and hard-to-scale issue of fixing failures in conversational service agents, though it appears incremental as it builds on existing workflow and retrieval methods.
The paper tackles the problem of LLM-based agents being error-prone and inconsistent in user-facing services by proposing WISE-Flow, a workflow-centric framework that converts historical interactions into reusable procedural experience, resulting in consistent improvements across base models in experiments on ToolSandbox and τ²-bench.
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 $τ^2$-bench show consistent improvement across base models.