AI-Enhanced Business Process Automation: A Case Study in the Insurance Domain Using Object-Centric Process Mining (2504.17295v1)

Abstract: Recent advancements in AI, particularly LLMs, have enhanced organizations' ability to reengineer business processes by automating knowledge-intensive tasks. This automation drives digital transformation, often through gradual transitions that improve process efficiency and effectiveness. To fully assess the impact of such automation, a data-driven analysis approach is needed - one that examines how traditional and AI-enhanced process variants coexist during this transition. Object-Centric Process Mining (OCPM) has emerged as a valuable method that enables such analysis, yet real-world case studies are still needed to demonstrate its applicability. This paper presents a case study from the insurance sector, where an LLM was deployed in production to automate the identification of claim parts, a task previously performed manually and identified as a bottleneck for scalability. To evaluate this transformation, we apply OCPM to assess the impact of AI-driven automation on process scalability. Our findings indicate that while LLMs significantly enhance operational capacity, they also introduce new process dynamics that require further refinement. This study also demonstrates the practical application of OCPM in a real-world setting, highlighting its advantages and limitations.

AI-Enhanced Business Process Automation: A Case Study in the Insurance Domain Using Object-Centric Process Mining

This paper presents an in-depth analysis of the transformative impact of AI on business process automation in the insurance sector, focusing specifically on the automation of claim part identification using Object-Centric Process Mining (OCPM). This case paper not only examines the intricacies of implementing LLMs to improve operational capacity but also assesses the broader implications of AI-driven transformations in business process management.

Introduction and Objectives

The paper begins by addressing the limitations of manual processes in handling insurance claims, particularly the identification of claim parts—a task identified as a key bottleneck due to the knowledge-intensive nature of the work. By leveraging LLMs, the paper argues that organizations can significantly enhance operational efficiency and scalability. To evaluate this transition, OCPM is employed as a methodological framework capable of analyzing concurrent traditional and AI-enhanced process variants.

Methodological Approach

The methodology adopted in this paper is structured around Business Process Reengineering (BPR) and CRISP-DM methodologies. These frameworks guide the systematic refinement and deployment of AI models within the insurance claim management workflow. The research outlines six distinct phases: understanding, initiation, model development, transformation, implementation, and evaluation. Data-driven strategies are pivotal throughout the phases to enable the successful deployment of AI, with particular emphasis on security, data quality, and alignment with business objectives.

Model development leverages techniques such as Chain-of-Thought prompting and few-shot learning, utilizing GPT-4o with Structured Outputs to automate claim part identification. The models were evaluated against human performance benchmarks, ensuring comprehensive understanding and mapping of business-related variables and linguistic factors.

Results and Analysis

One of the paper's significant contributions is its detailed analysis of process scalability observed through OCPM. AI-driven automation managed to substantially increase claim part identification from 1.82% to 27.62%, showcasing a remarkable scaling capability. However, this success led to new dynamics in the business process, shifting bottlenecks to the claim investigation phase, which now requires organizational adjustments to handle increased workloads.

Object-centric analysis within this paper, bolstered by the transformative potential of OCPM$^2$, facilitated nuanced decision-making by stakeholders. However, the paper also highlights challenges in visualizing and communicating process mining results, emphasizing the need for improved graphical tools and techniques suited for business stakeholder engagement.

Lessons Learned and Implications

The paper indicates several critical implications for AI-driven process automation:

• Realignment of Business Resources: AI interventions alleviating major bottlenecks could introduce new constraints that require strategic shifts in resource allocation and operational redesign.
• Holistic Process Evaluation: Successful AI implementation necessitates a comprehensive view of cross-process impacts, ensuring that scalability benefits are harmonized across the organization.
• Complexity in Visual Communication: Effective dissemination of process insights requires customizable and dynamic visualization methodologies to bridge technical understanding with business leadership.

Conclusion

The paper concludes that AI can transform business processes and operational capacities significantly, yet such advancements must be matched with adjustments in subsequent processes to realize full business value. The inherent complexity in OCPM visualization suggests opportunities for the development of better tools that can simplify communication between process modelers and stakeholders, thereby enhancing strategic decision-making.

Overall, this case paper provides valuable insights into the practical application of AI in business process management, presenting a framework for future research and development in AI-enhanced digital transformations across various sectors.