![enter image description here][1] **Research Proposal** The objective of ADA MINToring is to develop and evaluate an AI-based methodology that enhances learning efficacy and promotes gender equity in STEM education. This will be achieved by utilising a female-centric AI-based learning assistant, named Ada. **Research Questions** This study aims to investigate whether the utilisation of an AI-based learning assistant, namely ADA, in comparison to conventional differentiation measures, results in... *RQ1: an improvement in the learning effectiveness of mathematics lessons?* *RQ2: a reduction in gender-specific stereotypes?* **Pilot Study** The pilot study employs a quasi-experimental design comprising approximately 200 students, divided into an intervention and a control group. The intervention group interacts with the AI-based chatbot, Ada, whereas the control group employs the conventional differentiation technique, namely help cards. **Ethical Validity** The ethical validity of this research project has been subjected to a comprehensive evaluation and has been granted approval by the *Ethics Committee of the Faculty of Philosophy* of the University of Cologne. The committee's assessment has specifically confirmed the ethical soundness of the study's methodology, particularly with regard to informed consent procedures for pupils and their parents, as well as the appropriate handling of personal data through the implementation of anonymisation measures. [1]: https://files.osf.io/v1/resources/aqk5e/providers/osfstorage/672b70067f76ca1b77f5c975?mode=render