Cloudera has completed three acquisitions in roughly two years under CEO Charles Sansbury, each aimed at filling specific technology gaps in the company's platform for analytics and AI. The deals - for Verta, Octopai, and Taikun - reflect a deliberate build-versus-buy approach as enterprises grapple with deploying generative AI across fragmented data estates.
"All the acquisitions we've made are technology building blocks that we've repurposed to serve the strategy, which is to build out this platform for analytics and AI," Sansbury said at a media roundtable in Singapore.
The three acquisitions
Sansbury took the helm in August 2023, just months after OpenAI launched ChatGPT. The timing made one thing clear: corporate demand for AI development and deployment tools was accelerating. Cloudera's first move came in June 2024 with the purchase of Verta, an operational AI startup whose platform managed large language models across hybrid and multi-cloud environments.
By November 2024, the company had acquired Octopai, which provided automated data lineage and metadata management. Sansbury said customer conversations drove the decision. "We started hearing about the importance of data governance, data lineage, and data quality. We had some capabilities of our own, but we started looking for a data governance and data lineage tool."
The third acquisition, announced in August 2025, brought in Taikun, a cloud-native infrastructure platform for managing Kubernetes across hybrid and multi-cloud environments. Sansbury said this deal was different in scope. "It fundamentally changed how we think about the way the product is built and deployed." The company had considered using technology from large vendors but concluded it needed to own the containerisation layer as an embedded internal component, not to compete with companies like Red Hat.
Anywhere Cloud and the data problem
The Taikun technology became central to Anywhere Cloud, a platform Cloudera announced in August for building, deploying, and scaling data and AI applications across multi-cloud and on-premises environments. Sansbury said Taikun's orchestration layer was repurposed into the single operational control framework that underpins the platform.
The design responds to a persistent enterprise problem. Many companies attempted to consolidate proprietary data into one location to feed AI systems, only to discover that some data could not be moved, some should not be moved, and the data they did move proved expensive. "One of the founding principles of Anywhere Cloud is to let organisations leave data where it resides and carry out data engineering through an abstraction or control layer," Sansbury said.
Using external table technology such as Apache Iceberg, the platform creates virtual, federated data sets without requiring large-scale data re-engineering before work can begin. This approach brings compute to the data rather than forcing data migration - a practical answer to what Sansbury called "a messy data estate."
What comes next
Sansbury pointed to physical AI as an area under examination, alongside ongoing work to improve current generative AI use cases. He described the main enterprise applications so far as code completion, chatbot development, and next-generation customer service. "We've gained a great deal of productivity internally from these applications. I still think most companies are focused on such quick wins because these applications can produce results quickly."
On future acquisitions, Sansbury said the company has enough development work in progress for the next 12 months but remains open to targeted deals. The profile he described is precise: "It would be a small company with a product that has demonstrated market fit but hasn't achieved sales escape velocity, because then you obtain the product without having to pay for a revenue stream."
Why this matters for executives and strategy
Cloudera's M&A pattern offers a case study in using acquisitions to accelerate product roadmaps without buying revenue. Each deal targeted a specific technical capability - AI model management, data lineage, container orchestration - that the company could have built internally but chose to acquire for speed. For leaders shaping AI for Executives and Strategy, the approach underscores a practical reality: the difference between build and buy often comes down to time-to-market, not just cost. The data challenges Sansbury described - fragmented estates, unmovable data, expensive migration - are not unique to Cloudera's customers. They are the conditions most enterprises operate under, which makes the architecture decisions behind platforms like Anywhere Cloud worth watching for anyone responsible for AI Data Analysis Courses or enterprise data strategy.
Your membership also unlocks:
