AI agents already support a wide range of business functions, from procurement to IT. In this series, we’ll present real-world use cases and provide in-depth insights into the practical world of AI agents.
Part 1: Copilot in the procurement department of an international healthcare company
At the procurement department of a global healthcare enterprise with around 100,000 employees, many processes such as approvals, supplier inquiries, and document management are now supported by a Copilot Agent. Developed in partnership with Campana & Schott, the agent automates recurring requests, delivers information faster, and significantly reduces the workload of the procurement team.
The starting point: High effort for simple information requests
Procurement staff received numerous inquiries every day about orders, delivery status, and invoices. Each request required manual searches in the SAP system, followed by communication via email, chat, or phone—a time-consuming process that tied up valuable resources.
In the highly regulated healthcare sector, where speed, accuracy, and traceability are essential, this effort became increasingly unsustainable.
To relieve the procurement team and free up capacity for value-added activities, the company decided to intelligently automate frequent standard requests using an AI agent.
The solution: A Copilot Agent for procurement
Together with Campana & Schott, the company developed a Copilot Agent specifically tailored to procurement needs. The goal is to make information about orders and invoices available quickly and transparently via natural language without switching systems.
Built on the Microsoft 365 Copilot framework and integrated with SAP via Azure components, the agent accesses live operational data. Responses are delivered in natural language, enhanced with adaptive cards that clearly present relevant information. The focus is on combining intuitive usability with technical precision: responses appear as text suggestions backed by original SAP data, enabling users to confidently verify results.
The rollout began with a structured pilot phase and a defined user group. Campana & Schott supported both the technical implementation and ongoing optimization based on user feedback. The result: a solid foundation for scaling to additional departments and processes.
The value: Less effort, greater efficiency
Even during the pilot phase, the Copilot Agent's potential became clear. Standard inquiries are now resolved in seconds with no need for complex interfaces or system switching.
Thanks to adaptive cards, all delivered information is fully traceable to its SAP source. In a regulated environment like healthcare, this transparency is a significant advantage.
Beyond significantly reducing processing times, the agent frees the procurement team from routine tasks, creating space for strategic work while improving response quality and reliability.
The technical architecture was designed from the outset to be scalable for future use cases. In the long run, the agent will support other departments as well, such as key users who regularly handle process-related queries. This paves the way for company-wide relief from standard procurement inquiries.
At a glance
What the procurement Copilot Agent delivers:
- Noticeable relief for the procurement team through automated information retrieval
- Direct access to SAP data via natural language
- Transparent display using adaptive cards
- Fast response times and high data reliability
- Scalable architecture for future use cases
- End-to-end support from Campana & Schott—from idea to implementation
About AI Agents:
AI agents are transforming the way businesses operate. Intelligent agents now provide instant access to information and smoother workflows, replacing manual searches, tedious coordination, and system silos.
Employees no longer need to switch between systems or navigate cumbersome processes. Instead, they receive answers directly within their work context, whether they are working in procurement, sales, marketing, or service. This new type of process support increases efficiency, improves transparency, and enables better decision-making.
In the future, autonomous agents will also perform specific tasks on behalf of users. However, users will remain in control and be actively involved in key approval steps.
