Large Process Models: Business Process Management in the Age of Generative AI

WorkDifferentWithAI.com Academic Paper Alert!

Written by Timotheus Kampik, Christian Warmuth, Adrian Rebmann, Ron Agam, Lukas N. P. Egger, Andreas Gerber, Johannes Hoffart, Jonas Kolk, Philipp Herzig, Gero Decker, Han van der Aa, Artem Polyvyanyy, Stefanie Rinderle-Ma, Ingo Weber, Matthias Weidlich

Category: “Enterprise AI”

Article Section: Process optimization with enterprise AI

Publication Date: 2023-09-11

SEO Description: “Exploring Large Process Models’ potential in Business Process Management with Generative AI.”

Claude.ai Generated Paper Summary

Based on my analysis, this paper presents some novel ideas and innovations in applying large language models (LLMs) to business process management (BPM):

1. Proposes the concept of a Large Process Model (LPM) that integrates LLMs with traditional symbolic and statistical BPM methods. The LPM utilizes LLMs and potentially process execution data-specific foundation models, but also retains existing BPM tools and techniques. This allows combining the benefits of LLMs with the reliability and precision of symbolic methods.
2. Envisions the LPM as a system that can automatically generate insights and action recommendations for a specific process in a given context. The LPM would fuse process data, unstructured knowledge, and models across organizations to provide tailored, context-specific support. This could reduce effort and expertise needed for BPM.
3. Discusses using the LPM to increase process observability by extracting knowledge from unstructured sources and enabling foundation model based analysis without organization-specific training. This could facilitate discovering relevant data sources and conducting predictive analyses.
4. Proposes the LPM can enable continuous automated improvement by linking insights to actions and assessments to refine recommendations over time. This could converge process design, execution, and analysis.
5. Provides a feasibility assessment of LPM capabilities, arguing simpler applications like knowledge generation are achievable now but full automation of the BPM lifecycle poses open challenges.
6. Identifies risks like ethics, reliability, and costs that require extensive research and validation before large-scale LPM deployment.
7. Situates the LPM vision in relation to emerging works on LLMs for BPM, distinguishing its comprehensive scope and emphasis on integrating symbolic methods.

So in summary, the LPM represents an innovative vision for synergizing LLMs with existing BPM techniques to provide contextualized, automated support. The paper sets out an extensive research agenda while also recognizing open feasibility questions.

Keywords

Large Process Models, Business Process Management, Generative AI, Large Language Models, Knowledge-based systems

Author’s Abstract

The continued success of Large Language Models (LLMs) and other generative artificial intelligence approaches highlights the advantages that large information corpora can have over rigidly defined symbolic models, but also serves as a proof-point of the challenges that purely statistics-based approaches have in terms of safety and trustworthiness. As a framework for contextualizing the potential, as well as the limitations of LLMs and other foundation model-based technologies, we propose the concept of a Large Process Model (LPM) that combines the correlation power of LLMs with the analytical precision and reliability of knowledge-based systems and automated reasoning approaches. LPMs are envisioned to directly utilize the wealth of process management experience that experts have accumulated, as well as process performance data of organizations with diverse characteristics, e.g., regarding size, region, or industry. In this vision, the proposed LPM would allow organizations to receive context-specific (tailored) process and other business models, analytical deep-dives, and improvement recommendations. As such, they would allow to substantially decrease the time and effort required for business transformation, while also allowing for deeper, more impactful, and more actionable insights than previously possible. We argue that implementing an LPM is feasible, but also highlight limitations and research challenges that need to be solved to implement particular aspects of the LPM vision.

Read the full article here