Project 1: Building AI Agents that Work in ERP Systems

Enterprise Resource Planning systems form the backbone of many organizational processes, integrating activities such as procurement, accounting, sales, and logistics. In practice, these systems are continuously used by employees who create, modify, and process transactions. Their interactions generate digital traces that shape organizational workflows and influence process performance. However, when studying digital processes in controlled environments, one fundamental challenge emerges: real organizational behavior is difficult to replicate without actual users interacting with the system.

Within the Process Innovation and Automation Lab, an experimental infrastructure has been established to simulate and analyze digital business processes. The lab currently operates an instance of the open-source ERP system ERPNext, which is highly extensible and accessible through various APIs. While the system infrastructure has been successfully deployed, the current setup remains largely static because there are no simulated users interacting with the system. As a result, realistic process dynamics, including user decisions, delays, and potential errors, are missing.

This project aims to address this limitation by exploring how agentic AI can be used to simulate employees interacting with ERP systems. By designing autonomous or semi-autonomous agents that perform realistic system interactions, the project seeks to create a dynamic environment in which business processes can be executed, observed, and analyzed. Such an environment would allow the lab to simulate organizational behavior and generate realistic event data for research and teaching purposes.

The goal of this project is to design and implement a prototype that simulates employee interactions within the ERPNext system using agentic AI techniques.

Key objectives include:

1. Analyze typical employee interactions within ERP systems and identify relevant process scenarios that should be simulated.
2. Design a concept for agent-based interaction with ERPNext, including possible use of APIs, automation workflows, and AI based decision logic.
3. Implement a prototype that enables automated agents to perform realistic system actions within a dedicated ERPNext instance.
4. Explore the integration of orchestration or automation tools such as n8n or other open source components to coordinate system interactions.
5. Evaluate whether the simulated interactions produce meaningful system logs and process traces that can be used for research in process mining and automation.

The resulting prototype should contribute to the further development of the Process Innovation and Automation Lab infrastructure.
 

References

• Davenport, T. H. (1998): Putting the enterprise into the enterprise system. In: Harvard business review 76 (4), S. 121–131.
• Robert Jacobs, F.; ‘Ted’ Weston, F. C. (2007): Enterprise resource planning (ERP)—A brief history. In: J of Ops Management 25 (2), S. 357–363. DOI: 10.1016/j.jom.2006.11.005
• Claude (Hg.) (2026): The 2026 State of AIAgents Report. How enterprises are building anddeploying AI in production. https://cdn.sanity.io/files/4zrzovbb/website/cd77281ebc251e6b860543d8943ede8d06c4ef50.pdf,
• Anthropic (Hg.) (2025): How we built our multi-agent research system. https://www.anthropic.com/engineering/multi-agent-research-system?
• Anthropic (Hg.) (2024): Building Effective AI Agents \ Anthropic. www.anthropic.com/engineering/building-effective-agents

Project Requirements

• Development of a project plan and distribution of tasks among group members
• Analysis of ERPNext system interaction possibilities
• Design and implementation of an agent-based interaction prototype
• Integration of automation tools such as n8n or comparable open-source components if needed
• Documentation of the developed architecture and implementation
• Presentation of intermediate and final results

Prerequisites

• Students of Information Systems, Computer Science, or Software Engineering at bachelor or master level
• Interest in business process automation, ERP systems, and artificial intelligence
• Basic programming skills (for example Python, JavaScript, or similar)
• Motivation to experiment with open-source tools and system integrations

If you are interested in this porject, follow the steps below to submit your application.

1. Form a project group
Projects are typically carried out in groups of 3–5 students. We recommend forming a group with fellow students before applying.

2. Prepare your application
Send a short application including:

• your transcript of records
• a short motivation letter (about one page) explaining why your group is interested in the project.

3. Submit your application
Send your application via email to:mailto:jannis.nacke (at) icb.uni-due.de