AI technology is starting to work really well. Unfortunately, I’ve found that the management of machine learning code, data sets and models — and the integration of these into operational processes — falls well short of enterprise standards. This can create blockers to adoption and reduce successful outcomes, even in organizations that have adopted AI.
But organizations can take specific measures to mitigate the difficulties. I’ll identify some wish-list items that could improve things. First, though, let’s create an inventory of some challenges to making contemporary machine learning and data science operational.
Read the full article by Winnie Chang in Forbes.