I don't if this fits your parameters but it's the best visual explanation of how machine learning works:

http://www.r2d3.us/visual-intro-to-machine-learning-part-1/

Might give you some ideas.

Can you expand on what you mean by visually show the ML solution? Usually the ML results in a simplified result/classification. Are you looking for examples / diagrams of how it worked in a specific system? https://medium.com/@ageitgey/machine-learning-is-fun-80ea3ec3c471 is the best overview I've read on the subject. He gives an example of figuring out the price of homes.

One of Google's Directors, Aparna Chennapradaga, gave a high level overview for designing for machine intelligence.

Take a look: https://vimeo.com/181330432

It's not earth shattering but she lays a good set of principles to consider. Hope it helps :)

http://nerds.airbnb.com/designing-machine-learning-models/

More links inside the article.

Also, if you're at Hubspot you might be in Cambridge + able to come to this event: http://projects.iq.harvard.edu/harvarduxgroup/event/boston-chi-presents-googles-research-director-peter-norvig-chi-machine-learning

Peter Norvig is talking about designing for ML.

https://backchannel.com/an-exclusive-look-at-how-ai-and-machine-learning-work-at-apple-8dbfb131932b#.v40hnxy94

http://techblog.netflix.com/2015/04/learning-personalized-homepage.html

https://medium.com/@kwchang/product-building-for-relevance-66285faf65e2#.tbfcl4jn5

And then neither of these are really product design case studies, but may be useful nonetheless:

https://engineering.pinterest.com/blog/when-data-science-house-cards

http://www.r2d3.us/visual-intro-to-machine-learning-part-1/