How QuantSpark leveraged Generative AI and Machine Learning to streamline order processing, reduce manual input, and achieve significant cost savings for a leading wood-based panel producer.

The challenge

A leading producer of wood-based panels faced significant operational inefficiencies in order processing. The manual approach required deep subject matter expertise (SME) to match customer requests with order specifications before entry into SAP. This not only made the process labour-intensive but also created knowledge silos restricted to specific locations.

To address these challenges, QuantSpark transformed a feasibility study into a Minimum Viable Product (MVP), integrating an AI-driven solution to automate SAP customer order processing and enhance operational efficiency.

The solution

QuantSpark designed a robust AI and Machine Learning solution consisting of three key components:

1. Order Extraction:

   • Utilised multi-modal Large Language Models (LLMs) to accurately extract product and order details from unstructured customer orders, including PDFs.

2. Order Matching:

   • Applied codified business logic to structure extracted data and predict required product and order fields with high precision.

3. Iterative Improvement:

   • Compared AI-generated customer orders with SAP-confirmed orders to identify discrepancies, continuously refining accuracy through trend analysis and learning cycles.

The results:

€140k 

FTE cost savings achieved, with potential for further gains in full build

83%

Product matching accuracy achieved post-MVP stage

50k

Annual volume of orders processed automatically

Are manual order processes slowing your business down? QuantSpark’s AI solutions can streamline operations, reduce costs, and enhance accuracy. Contact us today to explore how we can transform your workflows.