What slows organizations down isn’t a lack of information, but the effort needed to make sense of it across countless documents. Recognizing this challenge, Axians developed DERA to help its customers fundamentally transform how they interact with their data. The objective is simple: provide the right information to the right person in the right format, without manual effort.
As Filipe Cerqueira, Innovation Lead, explains:
DERA doesn’t just save time. It changes how organizations use information. The challenge was already there. But with Large Language Models, we asked ourselves: do we evolve, or stay comfortable?
The team chose to evolve. What started as a proof of concept quickly demonstrated its value, delivering faster and more accurate results.
This shift was reinforced by real-world impact. “We tested the technology, and the results were really good,” says Inês Saraiva, Lead Data Scientist. “So, we started replacing what we had, and clients kept asking for more.” From that point on, DERA grew in direct response to client needs, continuously expanding its capabilities.
At its core, DERA removes the burden of document processing. Instead of manually searching through dense content, users can access relevant information instantly. “No one should have to read ten pages to find one key sentence,” Filipe notes. By leveraging advanced Artificial Intelligence (AI) techniques such as prompt engineering and Retrieval-Augmented Generation, DERA reads, interprets, and delivers information in a clear and actionable format.
Despite its automation capabilities, DERA is built around a human-centric approach. “We don’t want to remove humans from the process,” Inês explains. “AI supports decision-making, but validation must remain human.” This principle is essential, particularly in regulated environments where accuracy and accountability are critical. Filipe reinforces this perspective, emphasizing that semi-automation is key, with technology enhancing efficiency while people ensure reliability.
Over time, DERA has evolved into a comprehensive platform with four core capabilities. It enables information extraction from complex documents, supports content validation for compliance and quality control, organizes large volumes through document classification, and delivers concise insights through summarization. Together, these capabilities reduce manual effort, minimize errors, and allow teams to focus on higher-value tasks.
Looking ahead, DERA will continue to evolve alongside advancements in AI. “It’s a hopeful uncertainty,” says Inês. “New technologies will allow us to improve even further.” One of the next key developments is the move toward multimodal intelligence, enabling DERA to go beyond text and interpret images and more complex data types. This responds directly to emerging client needs, such as validating signatures or verifying the authenticity of visual content.
The long-term ambition remains clear. As Filipe concludes:
Our vision is a platform that transforms any document, in any format, into usable intelligence.