On Dec. 3, Cognite celebrated the grand opening of its new 30,000-square-foot global headquarters in Tempe. The Norwegian company plans to add more than 100 new employees to support its suite of AI to harness, streamline and translate industrial data for businesses.
“Cognite’s decision to establish its worldwide headquarters right here in Arizona is a clear endorsement of the unmatched innovation ecosystem that we have worked so hard here to build,” says Gov. Katie Hobbs during the ribbon-cutting ceremony. “What happens at Cognite is felt across the globe, and now Arizona gets to be a part of that.”
According to Cognite Board Member Alysa Taylor, the industrial AI business is valued at $20 billion today, with forcecasts estimating that number to swell to $90 billion over the next eight years.
“That’s what Cognite and the great state of Arizona has to look forward to,” she continues.
Tempe’s talent
Perched on the ninth floor of the Hayden Ferry Lakeside III, Cognite employees enjoy a panoramic view of Tempe’s distinct landmarks, including A Mountain, Mill Avenue and Tempe Town Lake. Beyond the picturesque scenery, the office’s proximity to ASU was a powerful draw for the company.
“Originally, Cognite wanted to be in Scottsdale,” explains Jonathan Keyser, founder and managing partner of Keyser Commercial Real Estate. “But they had heard about Tempe’s tech ecosystem, and after touring a few buildings, they made the decision to only look at options in Tempe.”
LEARN MORE: AZRE Forum will examine water, power and future growth
Considering AI demand is projected to proliferate in the coming years, having easy access to ASU and the talent it graduates was a primary driver for Cognite to locate in Sun Devil territory.
“At ASU, we’re proud to educate more than 30,000 engineers and more than 20,000 business students. These future-ready leaders are eager to contribute to the missions of companies like Cognite,” says Chris Howard, executive vice president and COO of ASU. “Tempe is the right place and Cognite is the right company. I’m confident we’ve only scratched the surface of what we can build together.”
Deciphering the data
With the popularity of dedicated AI agents and the accelerating ubiquity of AI across products such as web browsers, phone operating systems and social media platforms, Cognite’s value proposition may not seem unique. But these models are based on the written word, which poses a challenge for many businesses that want to use AI in their operations.
“Industrial data isn’t language — it’s temperature sensors, work orders and piping schematics,” explains Jason Schern, field CTO for Cognite. “That makes it difficult for AI to apply the same reasoning it typically does.”
To give AI the required context it needs to generate useful insights, Cognite identifies meaningful relationships that exist within the data, creating a framework of meaning the AI can reason from.
“Industrial AI is so important because many industries have grappled with siloed, unstructured data they couldn’t access to understand the health of their business,” Talyor adds.
For example, Schern points to a common process called root cause analysis (RCA). When a major incident occurs, companies examine why it happened and determine a solution to prevent it from occurring again.
“It’s a difficult undertaking that takes months even at the best companies,” he continues. “That means there are many incidents that never get investigated or mitigated because RCAs take so long.”
When working with a customer, Schern discovered that the most time-consuming part of their RCA procedure was gathering and analyzing the data. By using Cognite’s Data Fusion product to contextualize the data and its Atlas AI agent to provide insights based on those records in plain English, the process was shortened from six weeks to six hours.
“In reality, it only took about 30 minutes, but there’s a nice symmetry to ‘six weeks to six hours,’” he adds. “This is just one example of the hundreds of ways [these systems can be used] at a company. But it does help give a sense of what we mean by our moonshot, which is to drive $100 billion worth of value for our customers using Cognite technology.”
The phrase “time is money” is a common phrase in business for a reason, but in some industries, even a small hiccup equates to a major loss of revenue.
“Oil and gas companies lose something like $1,000 per second when operations are down,” explains Stein Danielsen, co-founder and chief solutions officer for Cognite. “Before AI, we’d use a query language to figure out what’s wrong [with a particular machine]. But that’s too difficult, especially if you’re losing $1,000 every second.”
Today, Cognite can compile readings and measurements together, then layer an AI agent over the top of the dataset that can answer questions from an operator. Danielsen compares it to having a conversation with a very knowledgeable person who can ask follow-up questions to determine what information is actually being requested.
Say, for example, a company has factories across the world that use a specific piece of equipment. If someone asks a general question about the performance of that machine, Danielsen notes that the AI agent will prompt the user.
“It might reply, ‘Which one do you mean?’” he continues. “Then, the operator can say, ‘Oh, the one at the Phoenix site,’ and then it will fetch the requested information. I can even talk to the agent in Norwegian, even if the documentation is in English since that doesn’t matter to AI.”
Keyser adds that bringing the headquarters of a significant international entity to Tempe is a huge win for the state — perhaps bigger than many realize.
“Cognite is the global leader in industrial AI, which is exactly the kind of business we want,” he concludes. “Of all the places in the world they could’ve picked, Arizona is where they landed. When you consider the number of companies within their ecosystem, Cognite’s decision to be here will have a lasting ripple effect.”
