DeepL targets AI translation for enterprises with new $100 million

Seeking AI language translation business customers, Cologne, Germany-based DeepL announced another funding increase valued at more than $100 million by public reports. Language translation is becoming an increasingly important function for companies operating in different regions and different demographics.

Basic translation capabilities have been available on services like Google Translate for decades. But the challenge was to enable more sophisticated translations for business use cases that capture not only the literal meaning, but also the right tone and context. This is an area where AI-powered language translation is starting to make an impact.

DeepL launched in 2017 and has been steadily improving its technology through deep neural networks. The new financing boosts the company’s valuation to more than $1 billion. The company has not publicly released the total raised.

“We are not disclosing this number, all we can say is that it is significant,” DeepL CEO, Jaroslaw “Jarek” Kutylowski told VentureBeat.

While Kutylowski is coy about what the actual funding amount is, he has very clear objectives about what the money will be used for. As the company grows, he noted, it will spend more on basic AI research, enter new product areas and also expand its portfolio to enterprise customers.

AI-powered translation is a growing trend

The earliest days of language translation were driven by basic pattern-matching techniques.

For example, a user types “hello” into a database that matches the equivalent in another language; for example French (“bonjour”). However, the basic semantic constructs of pattern matching are not scalable for larger-scale translations, where context and tone matter.

DeepL is one of many vendors that have applied advanced AI techniques to better translate human language. Google has improved its Google Translate service in recent years with a range of different approaches, including the use of a recurrent neural network (RNN).

Microsoft has been actively updating its Azure Translator service with AI models that the company claims improve overall quality. Meta (formerly known as Facebook) is also not left out of the party, announcing its AI-powered Universal Speech Translator (UST) project in October 2022.

Using a deep neural network approach to language translation

DeepL has developed a translation engine that relies on neural networks (NN) to derive accurate translations.

According to the company, it uses a new NN design to understand the nuanced interpretations of phrases and sentences and is able to convey them in a target language.

“We don’t disclose the details of our translation technology, but we can say that as a company we have always pushed the boundaries of how neural networks are designed to maximize translation quality,” Kutylowski explains.

The company’s original core vision was to break down language barriers – and Kutylowski stressed that the company continues to focus on this area.

“In the beginning, we understood that this vision was very much tied to specific translation,” he said. “As we continue to evolve as a company, we see us using the underlying technology to help people communicate in other ways as well – with new products that make communication easier.”

The ongoing business challenges of translation

There are plenty of challenges businesses face when dealing with translation that DeepL wants to help solve.

Kutylowski noted that the world is becoming more connected every year, which increases the importance of language translation and communication. With that growing demand, he says, localization approaches may be too slow and simply cannot be scaled.

“Tools like DeepL enable the end user, whether in the marketing, legal or other team, to communicate across borders and publish material directly, without the need to contact specialist teams or hire people. take,” said Kutylowski. “That creates completely new opportunities.”