Organizational Memory for Agentic Business Process Execution

2026-07-03

AI Summary

This work addresses the limitation of general-purpose large language models (LLMs) in capturing enterprise-specific process knowledge and the risk of knowledge fragmentation and inconsistency inherent in existing knowledge injection approaches. To overcome these challenges, the paper introduces the concept of “organizational memory”—a shared, governed layer for process knowledge that centrally manages dynamically evolving business rules and enables coordinated invocation by multiple agents. The proposed architecture integrates LLMs with knowledge extraction, retrieval-augmented generation, and process modeling techniques to support centralized governance and efficient knowledge evolution. Evaluation through a procurement scenario prototype demonstrates that this approach significantly improves the accuracy and consistency of agent execution as well as the efficiency of knowledge maintenance.

Abstract

LLM-based agents offer new opportunities for automating business process execution beyond the limits of rule-based systems. However, general-purpose LLMs lack the organization-specific knowledge required for reliable execution, which is typically fragmented across human-oriented artifacts such as policies, process models, and standard operating procedures. While such knowledge can technically be encoded in individual prompts or agent-specific retrieval setups, this approach does not scale in enterprises, as it gives rise to knowledge silos and rule duplicates, and makes consistent updates and learning across agents difficult. We argue that this calls for an organizational memory for agentic business process execution: a shared, governed, and agent-consumable reference layer of evolving organization-specific procedural knowledge about how work should be executed. We derive requirements for such a memory, propose an architecture for its curation and consumption, and demonstrate its effectiveness in a proof-of-concept based on a procurement scenario.

Problem

• organizational memory
• agentic business process execution
• LLM-based agents
• procedural knowledge
• knowledge silos

Innovation

• Organizational Memory
• Agentic Business Process Execution
• LLM-based Agents
• Knowledge Curation
• Process Automation