Causal reasoning is one of the main challenges in AI and a core task in many scientific and engineering disciplines. Accurate causal models enable robust behavior in Out-of-Distribution scenarios, which is essential for reliable inferences and Root-Cause-Analysis in real-world applications. However, traditional causal models are often computationally intractable, limiting their scalability to high-dimensional data and complex scenarios. To address these limitations, this master thesis will explore the combination of Large Language Model (LLM) agents with data-driven causal reasoning. The goal is to develop scalable and mathematically sound methods for Causal Machine Learning. During your thesis you will study and implement new scalable methods within Causal Machine Learning. You will collaborate with a global research team specialized in Causal Discovery, Causal Inference, and Root-Cause-Analysis. Ideally, your contribution will be part of a scientific publication and will have a real impact on Bosch use-cases.