Automating visual inspection tasks holds great potential in production environments. In this project, a demonstrator was developed that covers the whole end-to-end data pipeline. It includes image acquisition via a camera, the data connection via the Fraunhofer Edge Cloud, the analysis using a deep learning model and the visualization in a frontend for the operator. An exemplary use case was implemented, which aims to detect defects in the laser welding process of battery cells. In the future, more use cases can be integrated into the demonstrator due to its modular hardware and software design.