The conceptual design of complex products may be improved by applying knowledge available in existing designs (analogues), pre-defined components (parametric models), and functional models (rules). To represent and re-use this knowledge, AI techniques may be applied such as Case-Based Reasoning (CBR), Constraint-Based Modeling (CBM) and Rule-Based Reasoning (RBR). Case-Based Reasoning indexes a database with existing products with known performance characteristics, to suggest a first prototype solution that fits the requirements. The prototype can then be modelled and dimensioned with Constraint-Based Modeling, and its performance be verified with a network of functional relations derived by Rule-Based Reasoning and optimized with a numeric solver. In this way a first design concept can be quickly generated and evaluated as a first step towards a more elaborated design.