How NOV Built Data Confidence Across Eight Data Warehouses with Bigeye

NOV Inc., the "department store of the oil and gas industry," supplies cranes and drilling rigs to energy companies worldwide. Behind this global operation, Business Intelligence Operations Manager Adam Nilsen oversees eight different data warehouses, each representing years of technological evolution and business acquisitions.

By 2023, this complexity created a crisis: data issues took days to resolve, and Nilsen's team found themselves constantly defending their data rather than confidently presenting it.

"Over the years, going back 15, 20 years ago, we built data systems from the first iterations," Nilsen explains. "The technology has changed over and over again. Some of those older systems are still live but upgraded, so it's a hodgepodge of logic, a hodgepodge of tool sets."

NOV had merged four different business units, each bringing established processes. The result was functional but fragmented, with data quality issues often going unnoticed until they had already caused downstream problems. "We knew that we had holes in our data process," Nilsen admits.

When evaluating data observability platforms, NOV prioritized clarity over complexity. "When we sat down and I started going over the demonstration, it was the ease of use," Nilsen recalls. "How Bigeye is architected and how these issues are presented were far superior than any other tool at the time."

The validation came with NOV's first alert, NOV-001, which dealt with an accounts receivable issue. "Bigeye had it presented in clear text, exactly what was wrong. It gave the historical timeline, plus potential ways of how to fix it."

The transformation was immediate and measurable:

Shortened time to resolution: "We used to count this by days, and the metric itself has completely changed. Now we're able to resolve these issues before they get out of hand."

Column-level lineage: "Lineage was a game-changer, the idea of being able to see not just at a table level, but an individual column level, that there was a problem, and all the downstream consequences."

Stakeholder confidence: "For NOV to deliver the products that it needs to our customers, every presentation can't start with, 'Well, if the data's right.' If it does, we've lost."