Neural Networks to Check BOMs?

Ok, this may sound way out there and science-fictiony, but here's my question:

Suppose I have a bunch of data representing individual part numbers for the Bill of Material (BOM) for a product. If my task is to verify that the BOM has all the right parts, I would simply look at each piece of data and compare it against my own knowledge of the part's applicability to the product. I would know, for example, if the product is left-handed and green, there should be no right-handed or blue parts on the BOM. If I didn't know it was supposed to be green, but 9 out of 10 colored parts are green, I'd be suspicious of a blue part in the BOM. This begins to sound, to me, like decision making that could be handled by a neural network. And I'm sure that there are ways to implement a neural network in Excel. So, can anyone offer advice for further investigation? Can Excel be used to create a neural network that checks BOM data, if a large set of known good BOMs are provided for learning data?

This looks interesting: http://www.xlpert.com/Build_Neural_N...cel_sample.pdf