The organisation processes monthly invoices using a manual workflow. This process, which follows simple rules and requires no human judgement, involves repetitive clicking, copying, and entering data for each invoice. This manual process leads to prolonged processing times with limited scalability, the risk of human errors and the inefficient use of employee resources. Consequently, productivity is decreased, and bottlenecks emerge, suggesting that there must be an automated solution.

The thesis began with a systematic literature review of core frameworks and approaches to implement RPA as well as an exploration of the potential of Intelligent Process Automation (IPA). This was followed by a case study analysing and modelling the as-is process, defining requirements, and key performance indicators (KPI). Based on this and the literature, an optimised to-be process was created, and a proof of concept (PoC) was realised using Microsoft Power Automate Desktop. This PoC demonstrates the feasibility and that there are no technical restrictions for the actual implementation.

This thesis demonstrates that automating the invoice management process with Power Automate Desktop substantially reduces monthly runtime, effectively eliminating errors and enhancing reliability. The solution provides seamless scalability, accommodating increased volume without additional human effort. This also boosts stakeholder satisfaction by relieving finance personnel of monotonous duties. A successful Proof of Concept (PoC), utilising a substitute web application, confirmed all functional and technical requirements were met. Subsequent validation via a User Acceptance Testing (UAT) further confirmed its feasibility and the absence of technical barriers. However, a key limitation of Robotic Process Automation (RPA) solutions is their inherent dependency on underlying applications. Any updates to these will necessitate corresponding workflow modifications. Consequently, a robust maintenance strategy is crucial. With the to-be process model providing a precise blueprint for deployment within the live finance application, the project is now ready for implementation.