Lightup Data Inc., a startup that helps companies detect when their applications ingest erroneous or out-of-date data, has raised $9 million in early-stage funding.

The Series A round was co-led by Andreessen Horowitz and Newlands Ventures. In its announcement of the investment today, Lightup detailed that Spectrum 28 Capital, Shasta Ventures, Vela Partners and Incubate Fund participated as well. The startup has raised a total of $20.7 million to date.

“The funding will further accelerate and extend our development and delivery of even greater capabilities in the Lightup platform to streamline data quality processes and increase time-to-insight,” said co-founder and Chief Executive Officer Manu Bansal. 

Mountain View, California-based Lightup provides a platform for detecting issues in data pipelines. A data pipeline is a software workflow that automatically moves information from one application to another. Such workflows can, for example, stream information about product availability from a retailer’s inventory tracking system to its supply chain planning tool.

Malfunctions in a data pipeline can disrupt a company’s analytics initiatives. If a pipeline that supplies information for a revenue tracking dashboard starts delivering out-of-data sales data, the dashboard will become inaccurate. Similar issues can arise when datasets are delivered with missing or duplicate fields. 

To catch such errors, companies create so-called data quality checks. Those are algorithms that automatically scan the information flowing through a pipeline for potential errors. According to Lightup, its platform makes it possible to implement such checks in minutes rather than the months usually required.

It allows developers to create data quality checks through a no-code interface. According to the company, its platform works by determining what information usually flows through a data pipeline. It then spots records that break from the pattern and indicate an error may have occurred. 

If a pipeline usually processes 500 megabytes of data per day, Lightup’s platform can generate an alert when that volume drops to 300 megabytes. Such a decrease may suggest the pipeline failed to process some records.

The platform can detect changes in not only the volume of data that flows through a pipeline but also the phenomena the data describes.

A retailer might have a pipeline that moves sales logs to an analytics application for processing. According to Lightup, its platform could detect if the average value of the purchases those sales logs describe suddenly decreases. The platform can spot more subtle changes as well, as if the average purchase value suddenly starts increasing by 10% per week after two months in which it experienced only single-digit weekly growth.

Users can customize what type of data changes the platform detects and under what conditions. According to the startup, anomalous data points are detected by machine learning models built into its platform. Customers can train the models on their data to improve their accuracy.

The company says its platform is used by multiple Fortune 500 companies including McDonald’s Corp. and Gap Inc. Lightup claims to have grown its recurring revenue fivefold over the past year thanks to strong demand, though it didn’t provide more specifics. The $9 million funding round announced today will enable the startup to hire more workers, build new features for its platform and grow its partner ecosystem.

Image: Lightup