The pitch enterprises kept giving Cloudera for two years wasn’t really about innovation. It was about cost, and for a long time, nobody wanted to listen.
“To be quite direct, the cost discussion fell on deaf ears,” Cloudera CEO Charles Sansbury told CNBC’s Squawk Box Asia this week. “In the past year, the volume has gone up dramatically.” Companies moving AI workloads from testing into production, he said, kept getting blindsided by bills, especially for applications that need to run around the clock. A fraud detection system at a large bank, he added, can cost three or four times more on public cloud than on infrastructure the company already owns.
Cloudera built Anywhere Cloud to tackle exactly that problem. The company took the wraps off the platform this week at its Evolve 2026 conference in Singapore, pitching it as a modular system that stretches across public cloud, private data centers, sovereign infrastructure, and even air-gapped networks, all from one control plane, without requiring customers to relocate their data before anything works.
ExxonMobil has been testing it as a design partner for a few months. Anvesh Koripella, an enterprise architect at the company who has managed Cloudera deployments for a decade, said the first installation took roughly six hours of manual setup work. Running the same process last week took about 65 minutes, start to finish.
That speed matters for what ExxonMobil actually wants to do with it. Most of the company’s data streams in from sensors, gas meters, and analyzers scattered across refineries and oil wells worldwide, according to system architect Monika Thulasi. She said the goal is shifting away from batch-processed data toward real-time streaming, eventually using AI to catch equipment faults early and route alerts to engineers before problems escalate.
Cloudera isn’t alone chasing this shift. Databricks launched a similar sovereignty-focused system in June, though it still routes through Databricks’ own cloud rather than installing directly inside a customer’s environment the way Cloudera does.
Lian Jye Su, chief analyst at Omdia, said the timing fits Asia-Pacific particularly well, where companies are increasingly wary of losing control over data amid a complicated geopolitical landscape. He expects most large enterprises to eventually run both platforms rather than pick one.
