Case Study: NerdWallet achieves faster ML model training and lower costs with Amazon Web Services

NerdWallet Uses Machine Learning on AWS to Power Recommendations Platform

NerdWallet, a personal finance startup, needed to speed up its machine learning workflow because moving models from prototype to production was taking months and its data scientists were stuck with manual, inefficient environment management. The company turned to Amazon Web Services, using Amazon SageMaker along with existing AWS services like Amazon EC2 and Amazon ECS to modernize its ML platform.

With Amazon Web Services and Amazon SageMaker, NerdWallet streamlined model training and deployment, launched a TensorFlow-powered recommendations platform, and helped customers better match with financial products. The results were significant: model iteration time dropped from months to days, and training costs were reduced by around 75% even as the number of models trained increased.

NerdWallet

Sharadh Krishnamurthy

Staff Software Engineer