VAST Data, the AI Operating System company, today announced that DeepL, a global leader in Language AI, is deploying the VAST AI Operating System as a new data foundation for its AI research and development environments.

The deployment comes as DeepL continues to grow the capabilities of its Language AI platform across written and real-time voice translation. More than 200,000 business teams and millions of individuals now use DeepL to communicate, collaborate and operate across languages, increasing the scale and diversity of the AI workloads its research teams must support.

As those workloads evolve, the underlying data requirements are changing with them. Training increasingly sophisticated language models requires researchers to access and work with large datasets quickly and reliably, while newer areas such as real-time voice introduce significantly more data-intensive training requirements than text alone.

“DeepL’s researchers are constantly pushing Language AI into more demanding areas, and the infrastructure underneath them has to keep pace,” said Stefan Andresen, Chief Scientist, DeepL. “VAST gives us the performance and reliability we need while allowing us to make significantly more efficient use of the infrastructure we already operate. Just as importantly, it gives us a path to make data more easily accessible across our own data centres and, over time, cloud environments, reducing the manual data movement that can get in the way of research. That becomes particularly valuable as we continue investing in data-intensive areas such as DeepL Voice.”

DeepL is deploying VAST to provide a higher-performance, more resilient data environment while making more efficient use of the NVMe infrastructure it already operates. Based on current deployment plans, DeepL expects to achieve roughly three times the storage capacity from its existing hardware, giving research teams greater room to scale model development without a corresponding increase in physical infrastructure. DeepL will use VAST’s SyncEngine technology to migrate data from its existing environment onto the VAST platform, helping streamline the transition.

The VAST AI Operating System will provide a unified platform for file and object storage, giving DeepL greater flexibility to support evolving AI workloads while simplifying its underlying data structure.

Over time, DeepL also plans to extend its use of VAST beyond its own data centres and into cloud environments. The goal is to give research teams more consistent access to the data they need across infrastructure locations, reducing the manual copying of datasets between environments that can slow experimentation and model development.

This becomes increasingly important as DeepL invests further in real-time Voice translation. DeepL Voice, first introduced in November 2024, brought DeepL’s technology into live multilingual conversation, with recent advances enabling it to preserve characteristics of a speaker’s voice across languages. Developing and training these capabilities places significant demands on data access and infrastructure performance, making it critical to move efficiently between data training and production environments.

DeepL is a great example of how quickly the infrastructure requirements behind AI are changing,” said Jeff Denworth, Co-Founder, VAST Data. “As AI moves into production, infrastructure reliability becomes just as important as raw performance. Written language, real-time voice and the next generation of AI experiences all place different demands on data, but researchers should not have to redesign the underlying infrastructure every time the model or product changes. VAST AI OS was built to deliver that reliability at AI scale, giving teams a foundation they can trust, and then getting out of their way. For DeepL, that means the freedom to keep building whatever comes next.

VAST AI OS is designed to provide a common data foundation for AI environments spanning model development, training and inference, with industrial-grade reliability engineered for increasingly large and demanding AI workloads. Its architecture enables organisations to consolidate file and object data services while providing consistent, high-performance access to data across distributed AI infrastructure.