Industrial products and services are increasingly configured, planned and sold through online shops partially validating dependencies between basic systems and purchase options. Additional customer requests are described as flow text, not being processed automatically today and thus, requiring time-consuming back office support. InnoSale develops methods to increase the expressiveness of validation rules (rule-based expert systems) and to suggest relevant purchase options (case-based reasoning). Sales engineers will be supported in finding previous customer requests & orders and other suitable solutions quickly (natural language processing, semantic search) as well as identifying similarities between customers (evolutional clustering). Based on different company’s internal and external data sources, dynamic pricing algorithms are also one project scope (trend prediction, machine learning, artificial neural networks, Q-Learning). User experience will be improved supported by combining deep learning systems with augmented reality techniques, 3D modelling and 3D printing.