Databricks Inc. today announced that it has bought Row Zero Inc., a startup with a spreadsheet editor optimized for large datasets.
The acquisition price was not disclosed. Seattle-based Row Zero raised $13 million prior to the deal. Its backers included local startup funds and Wes McKinney, a prominent software engineer best known for his work on open-source data management tools.
Excel spreadsheets can store up to 1,048,576 rows, while Google Sheets supports about 10 times as much data per file. Row Zero, for its part, enables users to create spreadsheets with as many as 1 billion rows. That enables it to support a broader range of data analytics use cases.
The platform’s high data storage limits are made possible by the fact that it runs every spreadsheet in a dedicated Amazon EC2 instance. Row Zero co-founders Breck Fresen and Nick End held technical roles at the cloud giant before launching the startup in 2021. The use of a dedicated instance has the added benefit of speeding up large data processing workflows. Row Zero says it can filter, reorganize, summarize and visualize millions of rows near-instantaneously.
Workers who wish to perform more complicated data science tasks can use Python. They can write code from scratch or use Python’s analytics libraries. The list of supported modules includes pandas, one of the open-source projects created by Row Zero investor Wes McKinney.
Built-in connectors can copy records kept in cloud data platforms such as Databricks to a spreadsheet. When the original version of a record changes, Row Zero refreshes the spreadsheet. It also syncs data in the other direction. Users can copy the results of an analysis carried out in Row Zero back to the linked cloud data platform.
Those who don’t wish to perform analyses manually have access to a built-in artificial intelligence assistant. It’s designed to automate tasks such as summarizing and visualizing data.
Databricks says that the acquisition will help enhance Genie, the AI assistant built into its cloud data platform. The plan is to provide Genie users with a spreadsheet-based interface for processing records.
The assistant is available in three flavors. The first, Genie One, makes it possible to analyze datasets with natural language prompts. The second edition is called Genie Agents and enables companies to create customized versions of Genie One equipped with curated datasets. There’s also a developer-focused edition called Genie Code.
The new Row One-powered spreadsheet interface will be underpinned by three other Databricks products called Genie Ontology, Unity Catalo and Unity Gateway. The first tool helps the assistant interpret company-specific business terms. Unity Catalog and Unity Gateway, in turn, ensure that AI initiatives comply with an organization’s governance policies.
“As more finance, operations and go-to-market teams adopt Genie, incorporating a seamless spreadsheet experience is a must-have,” said Databricks co-founder and Vice President of Engineering Patrick Wendell.
Row One will continue to be available as a standalone product.
