Salesforce is open sourcing a tool that aims to make it easier to build scaled machine learning systems for enterprises, TransmogrifAI.
TransmogrifAI is a automated machine learning library for structured data that enables data teams to transform customer data into meaningful predictions, according to the company.
Salesforce explained it has been using TransmogrifAI to power its Einstein AI platform, but it wants to open up the project to empower other developers to build machine learning solutions at scale.
“In order to make machine learning truly work for our customers, we have to build and deploy thousands of personalized machine learning models trained on each individual customer’s data for every single use case,” Shubha Nabar, senior director of data science for Salesforce Einstein, wrote in a post.
Using this library, developers can automate data cleansing, feature engineering, and model selection all with only a few lines of code.
According to Salesforce, TransmogrifAI features the five main components of the machine learning process: feature inference, transmogrification, automated feature validation, automated model selection, and hyperparameter optimization.
Because of the library, Salesforce has been able to deploy thousands of machine learning models in production and reduced the average turnaround time for training performant models from weeks to hours.
“While this level of automation has been essential for us to scale for enterprise purposes, we believe that every business today has more machine learning use cases than it has data scientists, and automation is key to bringing the power of machine learning within reach,” Nabar wrote.