Hello, From my industrial experience I can say that Machine learning is always like an engine of the product. You need lot of wrappers on top of it e.g. data extractors, data providers and data exporters. You would need APIs to expose machine learning solution from engine dB to applications that can use results produced by the engine. The software architecture having machine learning engine would highly depend what kind of business application are you trying to solve.
"....depend what kind of business application are you trying to solve":
Just to illustrate the above valuable reply and noticing the trend being “costumer driven” product design, may I share the link of an institution http://www.cais.ntu.edu.sg/~pdm/index.htm where researchers are exploring machine learning approaches in that field?
It depends upon what aspect that you are really interested in the product design aspect. Unfortunately it is too generic and here are few that would give you some initial understanding of the interaction:
Sanchez, R., & Mahoney, J. T. (2002). Modularity, flexibility and knowledge management in product and organization design. Managing in the Modular Age: Architectures, Networks, and Organizations, 362.
Sim, S. K., & Duffy, A. H. (1998). A foundation for machine learning in design.AI EDAM, 12(2), 1
Wang, L., Shen, W., Xie, H., Neelamkavil, J., & Pardasani, A. (2002). Collaborative conceptual design—state of the art and future trends. Computer-Aided Design, 34(13), 981-996.
Monostori, L., Váncza, J., & Kumara, S. R. (2006). Agent-based systems for manufacturing. CIRP Annals-Manufacturing Technology, 55(2), 697-720.