NEW YORK – In 2014, Amdocs had more data than its systems could manage.
The multi-billion-dollar software and services provider was storing much of its data in on-premises Oracle databases. But at the time, the databases Oracle offered could not handle the volume of data Amdocs was suddenly ingesting. The cost and complexity of managing exponentially increasing data volume was rising.
Amdocs needed to find a data management platform built for big data. The Cloudera platform proved to be its data management salvation.
"We were dealing with huge data," Boaz Rubin, head of network software R&D at Amdocs, told TechTarget. "The main idea was to reduce the total cost of ownership (TCO). Oracle was very expensive, and the solution from Oracle did not scale like it does today -- they have improved with Exadata, but [at the time] they didn't have a good solution that could scale."
In addition to data volume beyond anything Amdocs had previously experienced, because the company provides technology products for communications providers that handle sensitive data such as personally identifiable information, Amdocs needed a data management platform with robust security and governance capabilities.
An untamable amount of data was a common problem at the height of the big data era. Improving internet connectivity resulted in new data sources -- mobile devices, social media platforms, online services, Internet of Things sensors -- that caused data volume to explode, but the systems used to prepare data and databases and data warehouses used to store data were built for a bygone era.
Companies such as Databricks and Snowflake were just emerging at the time. Though they've grown into trusted data and AI platform providers, their capabilities were nascent at the time. Cloudera, conversely, already had a six-year history by 2014, and its Hadoop-oriented HBase platform was built for large-scale workloads, so Amdocs chose to move from Oracle to Cloudera for its data management needs.
"There were not a lot of options like there are today, so we chose Cloudera," Rubin said. "There were some new companies, but nothing very robust. We needed a very robust solution."
A slow start, but then success
Transitioning from Oracle to Cloudera was complex, and it took time, according to Rubin.
Amdocs uses data management tools not only to harness and operationalize its own data but also to develop products it provides to its own customers. Amdocs' users did exacting work with what Amdocs provided through Oracle, managing PII and other sensitive data that if exposed could result in regulatory violations and stiff penalties. In addition,Rubin noted that Cloudera's platform is not as user friendly as some other platforms.
Amdocs is not alone among Cloudera customers hoping that the vendor's platform becomes easier to adopt.
Gainwell Technologies, which provides technology for the healthcare industry, similarly struggled to get started with Cloudera when it became a customer about six years ago. Even today, there is room to improve the user experience, according to Paddu Vedam, Gainwell's chief innovation and modernization officer.
"A way of getting Cloudera deployed much more easily, and [building] some of the front-end UI capabilities that Databricks has," he told TechTarget when asked what he'd like to see from Cloudera. "Cloudera has built a lot of those capabilities, but those are areas where they could [improve] so even business users could use Cloudera instead of just the development teams that are technically inclined."
Ultimately, it took two years for Amdocs to begin getting a return on its investment in Cloudera.
"The beginning was very hard," Rubin said. "To move from Oracle to Cloudera was very complicated. The customers did very strict work with Oracle."
Amdocs initially built a generic tool that enabled customers to store data as JSON files in Cloudera's database. Customers, however, wanted a product that enabled them to run SQL queries on their data, so they were not pleased. Amdocs subsequently built a tool that more closely mimicked what it had developed using Oracle, which satisfied customers.
"The telco market is very strict [regarding] TCO, so we were able to provide a very competitive solution with the right TCO thanks to Cloudera."
Boaz Rubin
Head of network software R&D, Amdocs
After the rocky start, Amdocs became more experienced with Cloudera, and what has evolved since has been a partnership that has enabled Amdocs to maintain control of exponentially growing data volume, operationalize increasingly complex data, and develop data products that serve the needs of its customers in the communications industry.
"Over the next five years, we built a robust solution that is able to build a report from the data, and that extracts the data," Rubin said.
In fact, when cloud computing became ubiquitous -- Amdocs' operations were on premises in 2014 – Amdocs became one of the first Cloudera customers to use Cloudera's platform in the cloud as a way to expand its storage capacities.
Over time, Amdocs has used Cloudera to build products that customers use to manage over 200 million subscribers.
"The telco market is very strict [regarding] TCO, so we were able to provide a very competitive solution with the right TCO thanks to Cloudera," Rubin said.
Recently, to see how much the change to Cloudera has ultimately benefitted Amdocs and its customers, Rubin compared Amdocs' TCO using Cloudera to what its TCO would be using database capabilities from vendors including Databricks, MongoDB, EnterpriseDB and DataStax.
"Some have fancier features that they tried to bring to me, but no one was even close to the right TCO that I'm looking for," he said.
Up next: agentic AI plans
While Amdocs has so far used Cloudera for its data management operations and to build data management tools for to its customers, Amdocs is now building agentic AI-powered tools for its users as well.
Like many enterprises, making sure that data used to inform agents is accurate was one of the main barriers Amdocs faced when building agentic AI capabilities, according to Rubin. In addition, predicting the cost of development and deployment was a challenge.
"When you need to satisfy different places, different clouds and different kinds of platforms, it can become very expensive," Rubin said.
By February 2026, Amdocs overcame the problems that initially prohibited it from launching agentic AI tools and delivered aOS, an AI-powered operating system for its telecommunications customers.
Given the regulatory and compliance concerns of telecommunications providers and the risks posed by agents accessing data in the cloud, Cloudera is not yet part of Amdocs' AI workflow. Instead, Amdocs is building and deploying AI, and enabling customers to use aOS, in its own controlled environment.
But with Cloudera recently introducing Cloudera Anywhere Cloud, a hybrid data and AI platform that enables customers to build and deploy agents and other AI applications not only in the cloud but also on premises and in hybrid environments, Amdocs can begin using Cloudera to build and manage AI.
"Only Cloudera provides a solution for on premises," Rubin said. "It seems Cloudera is going in the right direction by trying to be the bedrock of on premises. It's very exciting because a lot of [companies] are starting to realize that the cloud is very expensive and that tokens are very expensive. I believe that the whole world will come back to on premises more and more."
Eric Avidon is a senior reporter for Informa TechTarget and a journalist with more than three decades of experience. He covers analytics and data management.
