FRoSTA Achieves 60% Touchless Invoice Processing with SAP BTP and SAP Document AI

FRoSTA AG, one of Europe’s largest frozen food producers, has expanded its SAP landscape to automate invoice processing, resulting in about 60% of invoices received now processed automatically, and fewer errors from manual entry. 

The company, which processes roughly 30,000 invoices annually for goods received in Germany alone, historically relied on manual steps that slowed accounting and delayed approvals. 

FRoSTA modernized invoice processing without customizing SAP S/4HANA by building automation as a cloud extension on the SAP Business Technology Platform (SAP BTP). Using SAP Document AI to extract invoice data and SAP Build Process Automation to validate, route, and post invoices into SAP S/4HANA Cloud Private Edition, the company now runs a largely touchless booking process for paper, PDF, and Excel invoices that can scale over time. 

From Manual Workflows to AI-Driven Invoice Processing 

FRoSTA has used SAP BTP for several years, connecting SAP S/4HANA with SAP SuccessFactors and third-party tools, and enabling analytics via SAP Datasphere and SAP Analytics Cloud. This foundation allowed the company to implement invoice automation as a cloud-based extension rather than a core ERP modification. 

The first automation use case targeted invoice processing for goods received, where each document previously required at least three manual touches. Before automation, staff scanned incoming invoices, routed them for budget approval, and manually entered data into the ERP. Mobile limitations caused delays of up to two weeks when managers were away. 

Using SAP Build Process Automation and SAP Document AI, FRoSTA built a modular setup that connects directly to SAP S/4HANA Cloud Private Edition. Now, a monitored folder triggers an AI-enabled workflow that extracts, validates, routes, and posts invoice data with minimal human input. 

Fabian von Thun, Data and AI Manager, IT, FRoSTA AG, said, “The automation created using SAP Build Process Automation and SAP Document AI enabled us to increase the automation rate in our procure-to-pay process. Part of SAP Business Technology Platform, these solutions are convenient to use, while the platform supports us in developing application extensions for our custom use cases and in achieving a clean core approach.” 

SAP Document AI uses large language models (LLMs) to identify key data including purchase order numbers, invoice amounts, and supplier names across document formats and languages. SAP Build Process Automation then validates the information against SAP and third-party systems, triggers approvals, and posts invoices to SAP S/4HANA Cloud Private Edition, all while following existing business rules. 

Realizing Efficiency Gains in Procure-to-Pay 

The automation reduced cycle times and errors, freeing accounting staff from repetitive work. About 60% of invoices for goods received are now processed automatically, with the remaining 40% routed for approval or exception handling. 

FRoSTA uses SAP Task Center to surface workflows in a single workspace and SAP Mobile Start for mobile approvals by roughly 80–100 cost-center leads. Supplier adoption is underway, with 30% already using the updated process, paving the way for higher automation rates. 

Expanding Automation Beyond Invoices 

Invoice processing is the first step in FRoSTA’s automation roadmap. Future plans include integrating XML-based e-invoicing via SAP Integration Suite and automating customer order creation, where around 100,000 orders are still manually entered annually. 

The company also plans to expand SAP Task Center as a unified workflow hub across SAP SuccessFactors, SAP S/4HANA Cloud, SAP Ariba, and SAP Concur, reinforcing a single-entry, work-centric experience for employees. 

What This Means for SAPinsiders 

Integration and platform strategy matter. FRoSTA shows how touchless invoice processing can be implemented without modifying the ERP core, using SAP Build Process Automation and SAP Document AI as extensions on SAP BTP. This aligns with 2025 SAPinsider research, where 57% of surveyed organizations report that the ability to integrate SAP and non–SAP systems remains a top priority.  

AI-powered document processing is maturing. FRoSTA’s use of LLM-powered SAP Document AI illustrates scalable, template-free extraction of diverse invoice formats. Organizations facing high variability in supplier documents are likely to adopt similar AI-based methods. SAPinsider benchmark research shows that 76% of organizations focus on increasing productivity by eliminating repetitive tasks, and 52% explicitly target reduced manual touchpoints to improve process quality, reinforcing the strategic impact of touchless automation.  

Unified workspaces and mobile approvals drive ROI. SAP Task Center and SAP Mobile Start reduced approval delays from two weeks to near real-time, demonstrating the value of centralized, mobile-first workflow design, and aligning with research indicating that integrated task hubs and mobile access are increasingly prioritized for digital work processes.