Bidding to break from the pack in a crowded market, data integration provider Talend SA today is introducing an algorithm that evaluates data trust based upon such factors as data quality, popularity and user-defined rankings.

Talend Trust Score instantly assesses the reliability of an organization’s data to help companies resolve data integrity issues across increasingly sprawling on-premises and cloud environments. Data assessed by Trust Score is indexed, optimized and certified for quality to help organizations accelerate their move to the cloud and comply with regulatory rules more reliably, the company said.

The score has a variety of uses, said Gray Hardell, senior manager of solutions marketing. “Executives reconciling different opinions can get a score of the health of their data,” she said. “Users can gain confidence that their data is trustworthy. Developers want to be sure they can get data that has the most confidence possible.”

The service is based on technology developed by Stitch Inc., a self-service data integration company that Talend acquired for $60 million in 2018. The feature will initially be included in the company’s subscription service beginning in the fourth quarter of the year at no additional charge, although Hardell said Talend is “looking at ways we can expand this because we believe every company needs it.”

Trust Score automatically indexes data sets as they are loaded to provide a picture of health before data is used. Data can be assessed in multicloud, on-premises and hybrid cloud environments, including cloud data warehouse and leading commercial data lakes, the company said.

Metrics used to calculate data trust include validity, popularity, completeness, discoverability and usage. Talend said its Data Fabric also automatically resolves data problems by recommending the appropriate transformations using machine learning-powered semantic awareness.

The service comes at a time when many organizations are wrestling with major disruption to their information governance practices brought about by the COVID-19 pandemic. Data that was once safely secured on corporate servers has been spread across an uncontrolled mesh of home PCs, cloud file shares and mobile devices.

Many companies do some form of data quality scoring but don’t incorporate the full range of options, Hardell said. “They’re not taking usage and discoverability into account,” she said. “You can have a perfect quality score, but that doesn’t mean people trust the data.”

The metric was developed by Talend based on input from customers and observations of how people rate the quality of data sets. “It will be a living algorithm that we can use to constantly improve the data set based on input from customers,” said Laurent Bride, Talend’s chief technology officer. “Trust Score is the entry point and then there’s a closed loop where you can continually improve that score.”

Image: Pixabay