Cologne, Germany-based DeepL has announced a substantial funding raise, estimated to be well over $100 million, as it sets its sights on targeting enterprise customers with AI-powered language translation solutions. Language translation has become increasingly vital for businesses operating globally and across diverse demographics.

While basic language translation services like Google Translate have existed for decades, the challenge lies in providing more advanced translation capabilities tailored to business needs. These advanced solutions must not only convey the literal meaning but also capture the appropriate tone and context—a space where AI-powered language translation is beginning to demonstrate its significance.

DeepL, established in 2017, has steadily improved its technology by harnessing deep neural networks. The recent funding round has propelled the company’s valuation to over $1 billion, although the exact amount raised has not been publicly disclosed.

DeepL’s CEO, Jaroslaw “Jarek” Kutylowski, refrained from revealing the precise funding figure but emphasized the clear objectives for utilizing the funds. As the company continues to expand, Kutylowski stated that the funds will be allocated towards advancing fundamental AI research, exploring new product offerings, and broadening its portfolio to cater to enterprise clients.

AI-powered translation is a growing trend

The early stages of language translation primarily relied on basic pattern matching methods.

For instance, when a user inputted “hello” into a database, it sought an equivalent match in another language, such as French (“bonjour”). However, the fundamental semantic principles of pattern matching proved inadequate for more extensive translation tasks where context and tone played significant roles.

In the realm of advanced AI-driven language translation, numerous providers have been making significant strides. Google, for instance, has been enhancing its Google Translate service with various approaches, including the implementation of a recurrent neural network (RNN).

Microsoft has been actively upgrading its Azure Translator service, incorporating AI models that they claim enhance overall translation quality. Meanwhile, Meta (formerly known as Facebook) is not lagging behind and made an announcement in October 2022 about its ambitious project, the AI-powered Universal Speech Translator (UST).

Taking a deep neural network approach to language translation

DeepL has pioneered a language translation engine hinging on the utilization of neural networks (NN) to deduce precise translations.

As per the company’s claims, they employ an innovative NN architecture to grasp the subtle nuances of phrases and sentences, enabling them to effectively convey these nuances in a target language.

While the intricate workings of their translation technology remain undisclosed, Kutylowski, a representative of the company, asserted that DeepL consistently pushes the boundaries in the design of neural networks to optimize translation quality.

The original core mission of DeepL was to dismantle language barriers, and Kutylowski stressed that this mission remains at the forefront of the company’s endeavors.

“In the early stages, we perceived this mission as closely intertwined with translation, specifically,” he stated. “As we evolve as a company, we envision harnessing the underlying technology to facilitate human communication in various other ways, introducing new products aimed at enhancing communication.”

The continuing enterprise challenges of translation

DeepL is committed to addressing the myriad challenges that enterprises encounter in the realm of translation. According to Kutylowski, as the world becomes increasingly interconnected each year, the significance of language translation and communication continues to rise. He believes that existing approaches to localization may be insufficiently swift and scalable to meet the growing demand.

Kutylowski further states, “Tools like DeepL empower end users across various teams, be it marketing, legal, or any other, to seamlessly communicate and disseminate content internationally, eliminating the need for external specialized teams or hiring. This, in turn, opens up entirely new possibilities.”