Lemvigh-Müller: Accelerating wholesale distribution with agentic AI and touchless procurement
Driving touchless decisions across supply flows
Lemvigh-Müller A/S, Denmark’s largest wholesaler of steel and technical equipment, needed a better way to manage thousands of supplier confirmations arriving as emails and PDFs. To read, compare, and act on supplier confirmations with touchless processing, improved governance, and higher-quality data, the company deployed a multi-agent workflow with an SAP Cloud ERP Private solution, SAP AI Launchpad, and SAP AI Core infrastructure.
| Industry | Region | Company Size | Partner |
| Wholesale distribution | Herlev, Denmark | 1,270 employees | NTT DATA Business Solutions Denmark A/S |
touchless processing, reducing manual workload at scale.
matching accuracy across supplier confirmations.
of suppliers onboarded within the first few weeks.
CIO, Lemvigh-Müller A/S
Relieving pressure in procurement workflows with agentic automation
Since its founding in 1846, Lemvigh-Müller A/S has built its business around precision and speed. Today, the company operates 25 stores, 4 central warehouses, and 6 business areas, supplying approximately 650,000 item numbers to more than 22,000 customers across Denmark.
At that scale, even minor changes to a confirmed quantity, delivery date, or price can ripple quickly through operations, impacting purchasing decisions, internal planning, and how customer-facing systems reflect product availability. To keep everything moving smoothly, buyers rely on clear, reliable supplier information arriving at exactly the right moment.
Each month, thousands of supplier confirmations come as emails and PDFs across nine shared inboxes spanning multiple product areas. From steel to electrical and plumbing, each supplier confirmation had to be opened, read, and understood, then carefully checked against the original purchase order in Lemvigh-Müller’s existing business systems. Buyers searched for discrepancies—such as partial deliveries, shifted dates, or price changes—often embedded within documents that were 10 pages or more.
This process required focus and experience. At the same time, much of the work involved manually locating key data points and comparing them line by line. With confirmations spread across multiple inboxes, gaining a complete view of incoming supplier confirmations took time and made it challenging to respond as quickly as the business demanded.
Lemvigh-Müller had already taken steps to improve this area through earlier automation initiatives. However, the complexity of unstructured data and the need for precise comparisons against records in existing business systems highlighted the opportunity for a more streamlined approach. The company needed a better way to give buyers the clarity and confidence to act more quickly while reducing repetitive validation and surfacing deviations earlier.
Head of SAP ERP, Lemvigh-Müller A/S
Activating agentic AI to streamline procurement at scale
To move faster without adding complexity, Lemvigh-Müller introduced a new way of working built on an SAP Cloud ERP Private solution and AI Foundation. Three custom AI agents were designed to operate directly within the company’s procure-to-pay process. Instead of relying on manual review alone, Lemvigh-Müller uses AI agents to monitor incoming supplier confirmations, read emails and PDFs, and extract key data from unstructured documents.
What sets this approach to agentic AI apart is how the agents move beyond extraction. They interpret supplier responses, compare them with purchase orders in SAP Cloud ERP Private, and evaluate whether quantities, delivery dates, prices, or other commercial details align with what was ordered. When confirmations match expectations, the automated process continues. When deviations appear, the AI agents route cases into follow-up workflows or flag them for buyer review.
Developing the project with NTT DATA Business Solutions Denmark A/S, Lemvigh-Müller combined the AI agent design with a clean core strategy and close collaboration with business users. As shown in the diagram below, the agents are built on SAP Business Technology Platform, which provides a foundation to connect data, run agents, and scale the solution across the business. AI Foundation and SAP AI Core infrastructure, as part of the platform, support development and execution, while SAP Cloud SDK for AI enables AI agents to interact directly with SAP Cloud ERP Private and the company’s purchasing processes.
The AI agents use SAP AI Core to read and interpret emails and PDFs originating from Microsoft Office. They then compare this information with purchase orders in SAP Cloud ERP Private, as the system of record for quantities, delivery dates, and pricing. Using Cloud Foundry tools and standard APIs where available, agentic AI operates smoothly within daily procurement.
Freeing procurement buyers to focus where decisions matter most
With agentic AI operating inside its procurement processes, Lemvigh-Müller changed how work moves through the business. What once required careful, document-by-document review now flows faster and with far greater accuracy.
More than 90% of supplier confirmations are managed through touchless processing, surpassing the initial target of 80% to 85% within weeks of going live with the AI agents. With manual effort reduced by hours, routine validation tasks fade into the background of daily operations.
With approximately 98% matching accuracy with records in SAP Cloud ERP Private, buyers can rely on how confirmations are interpreted, compared, and routed. They no longer need to review every case. Instead, their attention shifts to the moments that truly require experience and judgment—when a delivery date changes, a price shifts, or a quantity doesn’t align. Everything else continues forward automatically, managed consistently and at scale.
Additional improvements have emerged in how quickly the business can see and respond to supply requests. Supplier confirmations are processed as they arrive, improving the speed and accuracy of how supplier data is handled and providing near-real-time visibility into purchasing status. Procurement teams can react sooner to supplier changes, adjust plans with better context, and communicate more clearly with customers about product availability and expected delivery.
Adoption has moved just as quickly. Approximately 23% of suppliers were onboarded within the first few weeks, signaling how rapidly the approach can take hold. With a scalable foundation in place, Lemvigh-Müller is well positioned to extend this model across additional supplier flows and into adjacent use cases.
Over time, the impact of agentic AI has become unmistakable. Less time is spent searching through documents and performing repetitive checks, while more time is spent making informed decisions, strengthening supplier relationships, and keeping operations moving with clarity and confidence.
Extending agentic AI across the flow of wholesale operations
With a working foundation in place, Lemvigh-Müller is focused on expanding the reach of agentic AI across the business. The initial rollout began with three of nine inboxes, and the next phase is to extend the solution across additional supplier flows to bring more of the confirmation process into the same coordinated, AI agent–assisted approach. As adoption grows, the aim is to create a more connected and consistent view of supplier commitments across categories and teams.
At the same time, this project is viewed as a starting point for Lemvigh-Müller. Additional AI-supported initiatives are already under consideration, and adjacent use cases such as invoice scanning are being explored alongside ongoing improvements to master data, process design, and the AI solution itself. Each step builds on what is already in place, making it easier to introduce new AI agents and extend automation into other parts of the procure-to-pay process.
The broader goal is to establish a repeatable model for applying agentic AI to complex, high-volume workflows. With SAP Cloud ERP Private and AI Foundation as the backbone, Lemvigh-Müller can continue evolving how work gets done—introducing new AI agents, refining existing ones, and scaling AI agent–based processes across the business.
Featured partner
NTT DATA Business Solutions Denmark A/S brought expertise in SAP Business Technology Platform, AI agent development, and a clean core approach to design and deliver the solution. Working closely with business users, NTT DATA enabled rapid iteration while building, integrating, and scaling agentic AI across Lemvigh-Müller’s procurement processes.