Awareness for the need to address ethical considerations in software development is ever-increasing. However, there is little clarity on how and when ethical analysis should be introduced into the development process. Using a case study approach, we argue that ethical analysis should be integrated at a very early stage of software development and requires interdisciplinary cooperation between technical and ethical specialists for it to be fruitful and cost efficient. In order to show this, we analyze the development process of an AI tool under development at the KNIGHT research project of Stuttgart University of Applied Sciences. Ethical challenges faced by the original plans of the tool, the interdisciplinary process of solving these problems, and the technical changes necessary to overcome them are detailed. The technical changes are fundamental and could not have been easily implemented at a later stage of development, highlighting the need for early-on integration of ethical analysis.