Towards Enforcing Company Policy Adherence in Agentic Workflows

TL;DR

This paper presents a modular framework for ensuring that LLM-based agents adhere to complex company policies by translating policies into verifiable guards that operate during agent workflows.

Contribution

It introduces a two-phase approach combining offline policy compilation into guard code and runtime enforcement, enhancing policy compliance in agentic workflows.

Findings

Successful demonstration on the $ au$-bench Airlines domain

Preliminary results show improved policy enforcement

Highlights key challenges for real-world deployment

Abstract

Large Language Model (LLM) agents hold promise for a flexible and scalable alternative to traditional business process automation, but struggle to reliably follow complex company policies. In this study we introduce a deterministic, transparent, and modular framework for enforcing business policy adherence in agentic workflows. Our method operates in two phases: (1) an offline buildtime stage that compiles policy documents into verifiable guard code associated with tool use, and (2) a runtime integration where these guards ensure compliance before each agent action. We demonstrate our approach on the challenging $\tau$-bench Airlines domain, showing encouraging preliminary results in policy enforcement, and further outline key challenges for real-world deployments.