Technology

Why Databricks Bought a Spreadsheet That Handles Millions of Rows

Martin HollowayPublished 2month ago3 min readBased on 5 sources
Reading level
Why Databricks Bought a Spreadsheet That Handles Millions of Rows
source:databricks.com

Databricks announced on September 24, 2026 that it has acquired Row Zero. Terms of the deal were not disclosed TechCrunch.

Row Zero was started by former AWS and Tableau engineers. It raised $10 million in May 2025 at an estimated $40 million valuation TechCrunch. It makes an online spreadsheet that can work with more than one million live rows.

Normal spreadsheets slow down or crash with that much data. People then have to use small samples or split data into separate saved files. Row Zero looks like a normal grid, but it stays linked to live company data that would normally need coding tools or special business software to use.

Databricks said in its official newsroom release that "Row Zero will combine the simplicity and flexibility of spreadsheets with built-in auditability, security, and governance" Databricks. Row Zero confirmed the deal separately on its own website.

Databricks closed $5 billion in funding in August 2026 and reported a $7 billion annualized revenue run rate as of September 2026 TechCrunch. CEO Ali Ghodsi said the company plans many more purchases like Row Zero in the future.

Row Zero follows other recent purchases. In March 2026, Databricks bought Quotient AI, which tests and trains AI assistants, and SiftD.ai, a very early interactive notebook startup. In June 2026, it bought Panther, a security startup for AI systems previously valued at $1.4 billion in 2021. In August 2026, it bought Electric, the maker of PGlite, a small and light database TechCrunch.

Databricks has set ways to find such companies. Databricks Ventures invests in innovative companies that share Databricks' view of the future for data, analytics and AI. The 2026 Built-On Databricks Startup Challenge was a global competition for early-stage B2B startups building core products on Databricks.

The broader context here is that most business planning still happens in spreadsheets. They are quick, flexible, and controlled by the person using them. They also sit outside the safety checks that track who can see data and where it came from. Companies have tried to fix this by forcing people into stricter tools or by copying data into saved sheets. The first removes flexibility and the second cuts the link to the original data, like working from a photocopy instead of looking through a window, so neither has worked well.

In my view, the bet is that safety checks can live inside the spreadsheet itself. The key questions are how live the link really is, what system runs the math, how typed changes go back to the main tables, and what happens to privacy and tracking when a finance worker shares a file outside the company. If Databricks gets those right without forcing people to rebuild their sheets, many more business users can work directly with trusted data. It is also worth flagging that young spreadsheet tools often struggle when many people edit at once or when AI writes formulas. Fixing that inside a safe system is harder than making a fast grid, and it will decide if this becomes an everyday tool.