Agent WARPP: Workflow Adherence via Runtime Parallel Personalization

TL;DR

WARPP is a training-free, modular framework that enhances workflow adherence in LLM-based task-oriented dialogue systems by dynamically personalizing and pruning execution paths at runtime, leading to improved accuracy and efficiency.

Contribution

Introduces WARPP, a novel runtime personalization framework that dynamically tailors workflow execution in LLM systems without additional training, improving adherence and efficiency.

Findings

WARPP outperforms non-personalized and ReAct baselines in accuracy.

Achieves larger gains as intent complexity increases.

Reduces average token usage across tasks.

Abstract

Large language models (LLMs) are increasingly applied in task-oriented dialogue (TOD) systems but often struggle with long, conditional workflows that involve external tool calls and depend on user-specific information. We present Workflow Adherence via Runtime Parallel Personalization, or WARPP, a training-free, modular framework that combines multi-agent orchestration with runtime personalization to improve workflow adherence in LLM-based systems. By dynamically pruning conditional branches based on user attributes, the framework reduces reasoning overhead and narrows tool selection at runtime. WARPP deploys a parallelized architecture where a dedicated Personalizer agent operates alongside modular, domain-specific agents to dynamically tailor execution paths in real time. The framework is evaluated across five representative user intents of varying complexity within three domains: banking, flights, and healthcare. Our evaluation leverages synthetic datasets and LLM-powered simulated users to test scenarios with conditional dependencies. Our results demonstrate that WARPP outperforms both the non-personalized method and the ReAct baseline, achieving increasingly larger gains in parameter fidelity and tool accuracy as intent complexity grows, while also reducing average token usage, without any additional training.