dotData, a provider of platforms for feature discovery, is releasing dotData Feature Factory, providing advanced functionality that empowers data scientists with a data-centric approach to feature engineering powered.
According to the company, dotData Feature Factory enables a paradigm shift in enterprise data solutions and will replace dotData Py, a Python-based data science automation engine first introduced in 2018.
“This new product provides our heart and core as an independent product,” said Ryohei Fujimaki, Ph.D., founder and CEO of dotData. “In past years, we have kept validating that Feature Discovery is the biggest pain in enterprise data solutions. The vision of the new dotData Feature Factory is to enhance all data solutions for enterprise organizations."
dotData Feature Factory automatically identifies and suggests feature spaces from enterprise data, including relational, transactional, and temporal data, allowing users to kick-start feature discovery and identify key signals from day one.
Feature Factory lets users programmatically define feature spaces and auto-generate 100X broader feature hypotheses using a data-centric approach to feature engineering that augments existing data and feature knowledge.
Further, dotData Feature Factory halts the process of re-inventing the wheel with an Analytic Database and Feature Descriptor that records every data and feature transformation step, giving data scientists the ability to capture data transformation know-how and build reusable feature discovery assets for themselves and their teams, according to the company.
The platform makes it easy for data scientists to build transparent, readable, maintainable, and easily scalable features that cover edge cases when processing features with new data, accelerating and simplifying the process of moving from experiments to production.
By generating feature spaces and discovering features in a data-centric and programmatic manner, Feature Factory offers enhanced collaboration, increased efficiency, improved model quality, reusability, reproducibility, scalability, and transparency.
According to the vendor, this innovative approach breaks down silos, enabling organizations to capitalize on the wealth of information available while improving the effectiveness of their downstream data solutions, including ML and AI.
For more information about this news, visit www.dotdata.com.