The KIBa project combines experimental battery material production with artificial intelligence to develop hybrid models that optimize the manufacturing of NMC cathode active materials (NMC622, NMC811) used in lithium-ion batteries. The motivation is twofold: Germany risks falling behind international competitors in digital production methods, and the growing demand for high-performance battery materials for electric vehicles and energy storage requires more efficient, scalable manufacturing processes. The project focuses on two core production steps: the precipitation/synthesis of NMC precursor materials, led by BASF, and the subsequent conditioning of calcined particles through grinding in stirred media mills, led by NETZSCH. Both processes are instrumented with inline sensors and analytics to continuously capture process and product data such as particle size distribution, pH, temperature, torque, and pressure. A large dataset is generated through systematic parameter studies, varying inputs like stirrer speed, temperature, precursor concentration, and grinding media. These experimental datasets are enriched with CFD and DEM simulations that derive physical quantities, such as mixing intensity and mechanical stress, which are difficult to measure directly but important for building transferable models. The modeling approach works in two stages. First, artificial neural networks are trained as black-box models on the experimental and simulation data. These networks then generate large volumes of virtual datasets, which serve as the basis for genetic programming. The genetic programming automatically selects and combines physical short-cut models to produce a semi-mechanistic hybrid model that can predict process-structure-property relationships with a high degree of physical interpretability. The final step is transferring these models to industrial-scale processes at BASF and NETZSCH, validating them against new data, and laying the groundwork for AI-driven process control. The project also includes an economic and ecological assessment based on measured energy and material flows. Partners include TU Braunschweig, BASF, NETZSCH, and Malvern Panalytical.