The potential impact of machine learning on businesses is undeniable, but has only reached a fraction of its potential. Difficulties around data usability, a lack of scalable models, and a shortage of engineers are constraining its impact on companies' bottom lines. Giants like Google and Facebook have taken the lead in AI adoption owing to the vast amount of data and talent available to them. Meanwhile, even within these organizations, trained models conduct a narrow set of tasks in tightly-controlled environments, and don’t adapt to new problems. Removing these barriers to machine learning adoption will help more businesses extract value from their data and expedite growth for existing users. Building on our experience at MIT, Google and Facebook, we are developing a new ways of deploying AI models - drastically reducing the required data and training time for each new model. To bring our technology to market, we are working with companies to proactively manage the long-term value of lithium-ion batteries.