AI Automation in Insurance: How LLMs Drive Digital Transformation

On April 24, 2025, researchers Shahrzad Khayatbashi, Viktor Sjölind, Anders Granåker, and Amin Jalali published AI-Enhanced Business Process Automation: A Case Study in the Insurance Domain Using Object-Centric Process Mining, exploring how AI-driven automation impacts process scalability in the insurance sector through a real-world application of OCPM.

Recent advancements in AI, particularly Large Language Models (LLMs), have enhanced business process automation, driving digital transformation. This paper presents a case study from the insurance sector where an LLM was deployed to automate claim part identification, previously a manual bottleneck. Using Object-Centric Process Mining (OCPM), researchers assessed the impact of AI-driven automation on scalability. Findings reveal that while LLMs significantly enhance operational capacity, they introduce new process dynamics requiring refinement. The study demonstrates OCPM’s practical application in evaluating AI integration, highlighting its advantages and limitations.

Recent advancements in process mining have transformed how businesses analyze and optimize their operations. At the forefront of this evolution is object-centric process mining (OCPM), a method that enables organizations to uncover inefficiencies and bottlenecks in complex business processes by tracking individual objects—such as products, orders, or customers—as they move through various stages.

Traditional process mining focuses on event logs, which record sequences of activities. However, OCPM goes further by incorporating object data, allowing for a more granular analysis of how different entities interact within a system. This approach is particularly valuable in scenarios where divergence (multiple paths) and convergence (merging paths) occur, such as in supply chains or customer service workflows.

How It Works

OCPM works by identifying patterns in event data that are tied to specific objects. For example, in a purchase-to-pay process, OCPM can track how an invoice moves through approval stages, highlighting delays or redundancies. By analyzing these interactions, businesses can pinpoint inefficiencies and implement targeted improvements.

A recent case study demonstrated the power of OCPM in identifying overstocking issues within a procurement process. By tracing individual purchase orders, researchers uncovered discrepancies between inventory levels and demand forecasts, leading to cost reductions and improved resource allocation. Similarly, in an after-sales service process, OCPM revealed bottlenecks in customer issue resolution, enabling faster problem-solving and higher customer satisfaction.

Challenges and Solutions

Despite its potential, the adoption of OCPM has been hindered by its complexity and the lack of standardized methodologies for implementation. To address these challenges, researchers have proposed integrating OCPM with artificial intelligence tools, such as natural language processing (NLP) and machine learning, to enhance pattern recognition and automate process optimization.

For instance, combining OCPM with AI-powered analytics can help businesses predict potential bottlenecks before they occur, enabling proactive decision-making. This integration not only simplifies the analysis of large datasets but also makes OCPM more accessible to organizations without extensive technical expertise.

The Future of Process Mining

As businesses increasingly seek ways to improve efficiency and reduce costs, OCPM is poised to play a critical role in driving innovation across industries. By providing deeper insights into complex processes, it empowers organizations to make data-driven decisions that enhance operational performance.

Moreover, the integration of OCPM with emerging technologies like AI and blockchain opens new possibilities for end-to-end process optimization. For example, blockchain can be used to create immutable records of object movements, while AI can analyze these records in real time to identify trends and anomalies.

In conclusion, object-centric process mining represents a significant leap forward in business process management. By leveraging this technology, organizations can unlock new levels of efficiency, transparency, and customer satisfaction—ultimately driving sustainable growth in an increasingly competitive landscape.