NEW YORK & MUNICH, April 08, 2025–(BUSINESS WIRE)–Celonis, the global leader in Process Mining and Process Intelligence, today shared the results of an innovative AI-driven solution that is optimizing Smurfit Westrock’s inventory management, unlocking significant value. Enabled by the Process Intelligence Graph and built on the Celonis Process Intelligence Platform, the AI Copilot is a prime example of how Process Intelligence empowers Artificial Intelligence to deliver tangible business impact.
Building on the impressive achievements, Smurfit Westrock is expanding its collaboration with Celonis to optimize additional core business processes across its global operations.
Unlocking AI’s Full Potential Through Process Intelligence
With operations spanning the globe, an inventory with hundreds of thousands of spare part numbers, and a series of recent acquisitions, Smurfit Westrock faced challenges in mastering data inconsistencies, such as duplicate materials, language inconsistencies, and synonym-based descriptions (e.g., “lubricant” vs. “oil”). These issues resulted in unnecessary inventory buildup, procurement inefficiencies (both upstream and downstream), and higher costs.
Using the Celonis Process Intelligence Graph and large language model (LLM)-based analysis, Smurfit Westrock was able to harmonize its master data, creating a foundation for AI-driven inventory management. The impact was immediate and measurable. Within just two months, Smurfit Westrock identified that many purchase orders were raised for spare parts that were already in stock. Additionally, a significant amount of spare parts had remained unused for over eight years. By leveraging the Celonis Process Intelligence platform, the company is now proactively cancelling unnecessary purchases for excess stock and optimizing stock utilization.
The data harmonization layer provided by Celonis also powers an AI Copilot that optimizes spare parts availability for plant engineers. The Copilot identifies the right materials based on real-time inventory, plant location, and purchasing patterns. Equipped with a natural language interface, the Copilot enables users to query inventory across global facilities in their own words—whether by technical descriptions or common industry terms. This eliminates the need for exact part numbers and simplifies the user experience.
“In August 2024, when one of our European plants had a breakdown, the supplier quoted us a significant delay for the supply of a replacement part,” said Brian Dodson, Business Process Improvement Manager and Celonis Center of Excellence lead at Smurfit Westrock. “With Celonis, we were able to find a plant in the region with the needed part and schedule a transfer order in under a day. This enables us to unlock new levels of efficiency, deliver enhanced value to our maintenance teams, and drive sustainability.”