Data management unicorn Collibra Inc. today disclosed that it has acquired a Maryland startup called OwlDQ, whose software platform uses machine learning to spot errors in companies’ business information.

The deal is the first acquisition for Collibra since its $112.5 million funding round last April. The New York-based company, which received a $2.3 billion valuation after the April round, sells software that helps large enterprises manage their business records. Organizations can use its technology to track where each individual file in their systems was created, who edited it and when, as well as set rules to prevent misuse.

OwlDQ focuses on a related task: helping companies clean up errors lurking their data. Data quality problems are inevitable in big organizations where a large number of workers interact with business records on a day-to-day basis. As a result, enterprises need to set up error filtering mechanisms to ensure that their analytics tools produce accurate insights and their customer-facing applications don’t accidently fetch incorrect information for users.

Those error filtering mechanisms usually take the form of handwritten database rules. OwlDQ replaces handwritten rules with machine learning models that, the startup says, catch issues more reliably and take far less time to set up.

The software can catch duplicate database entries, as well as identify cases where one copy of a record is different from another copy stored in a different location. The latter feature is useful for catching errors that arise as part of cloud migrations. When an application is moved off-premises, its information is copied from an on-premises system of record to a cloud system that often takes a different approach to organizing data, which can create technical challenges. 

OwlDQ also detects outliers in datasets. For example, if an e-commerce company’s product database contains an item with a list price that’s 100 times higher than what the average product costs, the software could determine that the price was most likely entered incorrectly. OwlDQ’s machine learning models perform the assessment while accounting for routine variances in data, such as seasonal price changes. 

“Collibra and OwlDQ will provide organizations with a single, cloud-based system of engagement to unify data governance, data privacy, data catalog, data lineage and now continuous data quality,” said Collibra Chief Product Officer Jim Cushman. The goal is to help teams “more easily and confidently get to trusted business insights and become data informed,” Cushman added.

The terms of the deal were not disclosed. Kirk Haslbeck, the founder and chief executive of OwlDQ, will join Collibra as vice president of engineering. 

Photo of Collibra CEO Felix Van De Maele: Index Ventures