Seven Months, One Interim Label
Debanjan Saha arrived at DataRobot in February 2022 and was interim CEO by July. Whether the permanent job produced a different company, the public record only partly answers.
In July 2022, the board of DataRobot gave its president and chief operating officer the whole company with one word attached: interim.
Debanjan Saha, now DataRobot's chief executive officer, had arrived in February. The board let the qualifier stand for about two months. On September 20, 2022, it dropped it.
The Interim Clock Runs Out
The company's own announcement supplies the sequence. Saha joined in February 2022 as president and chief operating officer, after twenty years leading multi-billion-dollar data and cloud businesses at Google and Amazon Web Services. He was named interim chief executive in July. By September 20 he was permanent, about seven months after he walked in. The release does not say what the board weighed in between.
The board chairman, Mark Hawkins, said the directors were confident Saha was the right leader to chart the path forward. Saha's own statement was shorter on biography and longer on ambition: "DataRobot is uniquely positioned to lead the AI industry with the most widely deployed and proven AI platform on the market". He promised to "execute with focus and agility".
The announcement offered one number to back the claim: DataRobot was trusted by customers including a third of the Fortune 50. Revenue, headcount and funding stayed private then. In every public source reviewed for this profile, they still do.
A Company That Renamed Itself
Read the boilerplate paragraphs of two press releases side by side. In September 2022, DataRobot described itself as the AI Cloud leader, selling a platform meant to bring together all data types, all users and all environments to deliver business insights. In June 2026, its releases describe the agent workforce platform, one that lets organizations create and scale AI agents that connect directly to business processes.
Between those two descriptions sits a four-year tenure and a change in what the company says it sells: from insights to agents. The documents do not say who made that call, or when. They show only that it was made.
Three Customers, One Argument
The argument surfaces three times in the first half of 2026, each time in Saha's words.
On January 26, Aon announced that it would use the DataRobot platform in client onboarding and servicing, including certificate generation, invoice processing and ID card issuance. Aon described the work as a collaboration to explore improvements, with colleagues keeping oversight. Saha framed it larger: "The insurance industry is rethinking how work gets done — not just to eliminate manual friction, but to deliver a better, more personalized customer experience".
On March 17, alongside Dell and NVIDIA, the pitch turned to control. DataRobot's release put the problem in numbers: assembling a platform for work off the big cloud providers can mean integrating more than 50 disparate tools, and, citing IDC research, 96% of organizations report higher-than-expected costs when deploying generative AI. Saha's answer was aimed at the cloud providers themselves: agentic AI that "answers to the boardroom". His reasoning: "The most sensitive enterprise data remains on-prem, and modern infrastructure spans edge to private to hybrid cloud." The release lists the uses the blueprint targets: real-time portfolio stress tests and fraud detection for finance teams, shipment rerouting when supply disruptions are detected, and patient triage in healthcare.
On June 2, the setting was an oil major. Each of Chevron's robotic inspection missions, the release explains, required operator verification of operating conditions through a permitting process, adding time and steps before work could begin. DataRobot's agents would assess conditions continuously instead. Chevron uses aerial and terrestrial robots worldwide to inspect equipment for abnormal conditions. The release says the system applies a Safe Start agentic assessment, running on NVIDIA microservices inside DataRobot's platform, and draws on existing hard-wired gas sensors, supplemental vision systems and AI models to judge conditions at the asset. Saha: "Agentic AI enables a shift from manually approving each mission to continuously assessing conditions in real time." Cari Armpriester, who runs Chevron's Facilities and Operations of the Future program, said the approach cuts reliance on manual, point-in-time checks and lets Chevron deploy robotics more efficiently.
Announcements, Not Yet Outcomes
Here the record thins. Aon's release speaks of intentions: plans to deploy, aims to accelerate. The Dell release announces a blueprint available through the Dell Automation Platform. The Chevron release describes an approach being applied. None of the three reports a measured result: no hours saved, no cost avoided, no contract value.
For a chief executive in his fourth year, that is the open question. The sources reviewed show a company that has changed what it says it sells and found large customers willing to be named. They do not yet show what it has earned from either.
A Seat Outside The Company
On September 30, 2026, the board of Vertex, Inc. elected Saha an independent Class II director, effective October 7, for a term expiring at its 2028 annual meeting. The SEC filing describes him as DataRobot's chief executive since September 2022, a former vice president and general manager of data analytics at Google Cloud and of database services at Amazon Web Services, and a holder of a PhD in computer science from the University of Maryland.
In 2022, one board removed a qualifier. In 2026, another gave him a seat beyond the company he runs. The next test is the one the releases leave blank: the first result, in a customer's own numbers, from an agent that DataRobot built.