Mariusz Pala, CTO of Generis, a  content, data and business process management company, is frank about the growing pains his team faced when scaling data management capabilities for heavily regulated industries. 

For starters, getting search right was a real pain point, Pala tells The Stack.

Generis Enterprise Technology, founded in 1997 and headquartered in the UK, is the creator of the CARA Platform. It serves over 800,000 users globally, across life sciences, financial services, energy, engineering, the public sector, and other regulated industries. 

Its high-profile customers include the likes of Bayer, Fannie Mae, Merck, Pfizer, and Visa. 

Customers use CARA for the likes of managing, categorising, and indexing content for compliance (for example, to organise clinical study reports, or health authority correspondence), and meeting industry data and content standards. 

Generis released the original CARA Platform in 2002. In 2019 it launched its cloud-first, next-gen CARA Platform, and in 2021, the CARA Life Sciences Platform. Among other capabilities, the latter helps pharmaceutical firms ensure traceability and transparency across their R&D workflow. Generis’s modern platforms are powered by Elasticsearch.

Billions of complex documents? Search got gnarly

Generis’s enterprise customers increasingly have to store millions, and in some cases even billions, of documents. They need to deploy capabilities like granular version control, content lifecycle management and advanced watermarks that keep them on the right side of regulatory standards.

With the old version of CARA, finding information was becoming harder to achieve quickly, as document and data volumes continue to grow. 

Organisations therefore needed the technology to transform complex processes, help them to work with documents and records from their creation to archiving, comply with regulations, index and find them fast. 

RDMS with separate Full-Text engine: Not a fast combination!

The initial release of Generis’s CARA Platform came back in 2002. 

As CTO Mariusz Pala recalls, when CARA was first created, the search experience was poor; it could take up to 30 seconds to complete a document search. 

This early version of CARA acted simply as a front-end for a different content management platform and a relational database from another vendor.

One of the main drivers when creating the new CARA Platform was dramatically improving the search performance of the system, particularly as previously all the data was stored in two databases. 

“You therefore had to search for whatever you wanted in one database and then complete a search in the second before merging the results, while applying the necessary permissions,” says Pala. 

In the early 2000s end users accepted this experience as there wasn’t an alternative option on the market. But the company realised it needed to deliver a dramatically improved experience: “Basically, we realised we had to build our own platform from scratch to solve all the issues and ensure a better user experience for our customers,” he adds.

(Its users now include eight of the 10 largest Life Science companies in the world.)

Generis decided to move away from the previous back-end platform and build CARA with its own content store. The inspiration came to them through a close collaboration with one of their customers. 

Sacramento introduces Elastic …

The City of Sacramento in the United States, a Generis customer since 2012, had a large volume of documents stored in CARA and needed to make them accessible to different functions within the company.  

Sacramento were already publishing documents onto an Elastic database to make them public for citizens every night; this showed Generis the technology could handle the traffic, and provide fast search. 

“We were on AWS, as there was no Elastic Cloud [at the time]. [But then] OpenSearch launched 2.0 which was not backwards compatible… it meant we would have had to rewrite the software,” recalls Pala.

“So we said, ‘No, thank you’!”

Elastic Cloud’s launch brought “excellent support”

With Elastic’s APIs being forward-compatible and the launch of Elastic Cloud, a route forward became much clearer. 

As CTO Pala explains “Widespread compatibility, robust APIs and excellent support from the Elastic team were all key factors in our decision to build ‘new’ CARA with Elastic.”

Having rebuilt for scale on Elastic, Generis has seen dramatic improvements in performance and impressive growth. The new CARA Platform offered a new approach to managing data and documents centrally, it allowed customers to replace a myriad of disjointed systems with one single platform. 

“Initially we were worried that customers wouldn’t want to move away from known solutions, but with the support of the Elastic team, six years on, 95% of our customers have chosen to upgrade to the CARA Cloud Platform.We now have more than 800,000 users who rely on the CARA Platform, and to keep up with the demand, our workforce has quadrupled,” he says proudly.

The majority of CARA’s users now deploy via Elastic Cloud (hosted on AWS).

And Generis currently maintains a capacity of 21 terabytes, distributed across three nodes with seven terabytes each –  returning blisteringly fast search results even in complex environments. 

One of the challenges is managing the variety of client requirements.  “You can have a billion documents or 20 million documents, with 500 million audit trails… One document with index content is like 30 to 50KB; one customer has one-page PDFs, another one has documents with 200,000 pages. So, content volumes differ from one kilobyte to 100 or 200GB.” comments Pala.

In order to reduce any downtime when re-indexing content for faster search with Elastic, the Generis team works around the clock. It also anticipates that the new Elastic Serverless offering will enable it to back up, restore, and create the data migration path from the existing cluster solution. 

The August 2024 launch of Elastic 8.15 was also a notable moment, Pala says, adding that this let Generis developers implement AI search features with out-of-the-box semantic search and generative AI technology. 

Generis is on a path of constant innovation, the CARA AI Assistant was released in March 2025 and each software update introduces new functionality to improve workflows, access and performance. 

The partnership with Elastic allows them to reach unprecedented levels of  speed, stability and security. 

Delivered in partnership with Elastic.

The link has been copied!