Machine learning and artificial intelligence (AI) are expected to drive a competitive advantage across most industries. Spending on these technologies has increased rapidly, yet the solutions are not realizing their full potential. Part of the problem is that organizations often lack the expertise to deploy these solutions at scale. Other barriers are a proliferation of tools, technologies, data silos, and "one-pipeline-per-machine-learning-process" thinking. Teradata's Analytics Strategy can solve these issues and enable organizations to get greater value from their data and advanced analytic solutions. The strategy notes that organizations can only successfully scale their machine learning and AI initiatives if they pay greater attention to feature reuse and model deployment. A second part of the strategy requires feature engineering and model scoring to be directly aligned with Teradata's core value propositions.
Ubiquitous machine learning will see organizations deploying hundreds of millions of predictive models in production. Teradata Vantage™ has the ability to scale machine learning vertically by training models on more than 1 million observations, and scoring them against more than 250 million observations multiple times per day. It scales horizontally by training millions of predictive models to support so-called "hyper-segmentation" use cases and scoring them daily in demanding production settings.