Lume AI Inc., a startup that helps companies move data between their applications, has raised $4.2 million in seed funding to enhance its technology. 

General Catalyst led the round. Lume said its funding announcement today that Khosla Ventures, Floodgate, Y Combinator and Soma Capital chipped in as well along with a number of angel investors.

Moving data from one application to another often requires changing its format. For example, such modifications might be necessary if the source application stores each record in a separate database row while the target system spreads records across multiple rows. Developers usually don’t change data formatting manually, but rather use scripts that can reformat thousands of records or more at once.

Creating data mapping pipelines, as such scripts are known, can require a significant amount of manual work. One reason is that companies must develop a large number of pipelines to cover all their data assets. Additionally, those scripts have to be actively maintained over time to avoid technical issues. 

Lume has developed a software platform that promises to ease the task. According to the company, developers can upload a dataset and specify how they wish to reformat it before loading it into a new application. From there, it automatically generates a script capable of performing the necessary changes. 

“The key foundation lies in Lume’s semantic understanding of data, allowing our AI to understand the nuances between data systems and create the mappings between them,” co-founder and Chief Executive Officer Nicolas Machado detailed in a blog post. 

The company also promises to simplify a number of related tasks. Lume’s platform can detect if the data it processes contains errors or other technical issues, such as missing fields. When developers change a data mapping pipeline, the software automatically makes adjustments to prevent the update from disrupting the flow of information. 

Lume’s data mapping features support JSON, a file format commonly used to move records between applications. There’s also beta support for CSV spreadsheets and XML, a data format that has a similar purpose as JSON but can support more complicated records.

One use case that Lume is targeting with its platform is processing customer data. A startup that offers a cloud-based data visualization tool, for example, might require users to organize their records into a certain format to optimize processing. Lume can automate some of the manual work involved in the process. 

The platform also makes it easier for companies to move data between their internal systems. In the enterprise, analytics projects often draw on information from several internal applications that organize records in different ways. The platform can automatically harmonize those records into a single format. 

The company is promising significant time savings for developers. According to Lume, one analytics provider compressed a four-week data mapping workflow into four days with its software. Another customer uses the platform to power about 1,500 automated data management workflows.

The seed round announced today will enable Lume to recruit more data management and artificial intelligence experts. With the new employees’ help, the company plans to roll out an array of enhancements for its platform. It’s developing features that will help companies test their data management workflows for errors and collect records from customers.

Image: Lume