ZyloTech, a MIT spin-off launched in 2014 as DataXylo, has announced its own solution to understand customers based on their data. It uses two – yes, two! – layers of machine learning to ingest, clean, unify, and analyze customer data from hundreds of sources; uncover intent; and deploy promotions. One part of the story is this happens continuously and in real time. Another is that unification and analysis skills are pre-trained on “vast sets of customer data”, which should greatly speed deployment and improve results. It’s another example of using current technology to overcome the bottlenecks that prevented traditional enterprise data warehouse projects from meeting marketers’ needs. ZyloTech doesn’t call itself a CDP but that shoe seems to fit.