Bayer Global Business Services cut manual errors in its procurement workflow by 70% and vendor/pricing errors by 60% with UiPath RPA plus SAP automation, shrinking purchase-order creation time from 10 minutes to 3 minutes per request and handling peak volumes of 900 requests per day across Brazil, Germany, and the USA, with 19 additional countries queued for expansion.
"Every automation starts with a need or a problem. People enter data by hand, which takes a long time and can cause mistakes." — Meritxell Corbeto González, RPA Project Manager, Bayer GBS (Barcelona office)
Procurement inside Bayer's Global Business Services had long been a bottleneck. GBS spans finance, HR, IT, and procurement across dozens of countries. In procurement alone, peak periods such as January and November generate 900 purchase requests per day, each requiring validation, vendor lookup, and data entry across multiple SAP systems. About 30% of all purchase requests bypassed any automation entirely, and 36% of those were for amounts under €5,000, meaning senior employees routinely spent time on small-value transactions.
The transformation moved in two phases. A UiPath robot was deployed at the front of the purchase-requisition-to-purchase-order workflow: when a request arrives, the robot logs into SAP, validates compliance rules, assigns vendor details, and generates the purchase order without human intervention. After the core flow stabilised, generative AI was layered on for exception handling. Both phases shipped within roughly a year, delivered by a team of eight with subject-matter experts providing process knowledge and User Acceptance Testing.
Architecturally, the solution combines UiPath RPA with SAP integration and a generative-AI layer for exception resolution. It sits inside Bayer GBS, which serves Bayer AG — a German multinational pharmaceutical and life sciences company headquartered in Leverkusen with €46.6 billion in annual sales and roughly 93,000 employees worldwide. Corbeto is based in the Barcelona GBS office.
The deployment operates at production scale across Brazil, Germany, and the USA, with 19 additional countries across South America, Europe, Africa, and Asia on the planned expansion roadmap. The target future state is one minute per request — a 3× reduction on the current 3-minute processing time.
Future direction is to roll out the existing automation to the additional 19 countries, push per-request processing time toward the one-minute goal, and apply the same UiPath-plus-generative-AI pattern to other GBS workflows — finance, HR, and IT. As Bayer's Spanish procurement manager Jose Carlos Esclapes Giner puts it, "The implementation of RPA for this process has significantly streamlined our operations, reducing the time spent on manual quality checks and allowing our team to concentrate on higher-value tasks, resulting in improved efficiency and accuracy."