How working across continents and time zones shaped Samsung’s European AI language development hub
As Samsung continues to lead in premium mobile AI experiences, we visited Samsung research centers around the world to see how Galaxy AI enables more users to reach their full potential. Galaxy AI now supports 16 languages, so more people can expand their language capabilities, even offline, with on-device translation features like Instant Translation, Instant Translator, Notes Assistant, and Navigation Assistant. But what does the evolution of AI language entail? Last time we visited India To learn how teams are working with students and universities to bring Galaxy AI to more people. This time, we’re in Poland to find out how European countries are working together to achieve their goal.
There is a saying at Samsung R&D Poland (SRPOL): “A day in SRPOL lasts 96 hours”. It points to its global role as one of the largest and fastest-growing R&D centers in the region, often operating across four different time zones. Located in the heart of Europe and covering many European and global markets, SRPOL has worked on automatic speech recognition, neural machine translation, and text-to-speech models for over 30 languages. When it came to integrating 10 languages into Galaxy AI, this experience meant the team was well prepared to seamlessly blend cultural perspectives with Samsung’s global technology.
SRPOL has years of experience in natural language processing. What makes it unique is its ability to adapt to work in any language thanks to its passionate team and tools, such as: Crowdsourcing Allowing for rapid and agile development.
“Collaborating across the continent means relentless data collection, feedback, and research, which has become something we really enjoy,” says Cornel Jankowski, head of speech decoding at SRPOL. “We’ve worked with so many languages that our team has developed global language-independent skills. When we’re asked to support a new language model, everyone’s attitude is: That’s great, let’s learn something else, it’ll be fun!».
European Centre for Artificial Intelligence Language Development
Language is a cornerstone of culture and communication across Europe, regardless of whether it is embedded in technology. However, it presents unique challenges for the SRPOL team, which is developing AI models for European languages.
“Every language and culture it’s part of has hurdles that make us reevaluate how we perceive a particular topic,” explains Adam Ross, head of AI at SRPOL. These hurdles include overcoming the untranslatability of certain phrases and dealing with idiomatic expressions that may not have direct equivalents in other languages.
The team saw these challenges as an opportunity to make SRPOL a European hub for AI language development. The biggest benefit of this is that it shortens the communication path between different departments, and essentially, the decision-making path. Whether it’s automatic speech recognition, neural machine translation, or text-to-speech, teams can simply communicate with their mobile QA colleagues and solve problems together efficiently.
While this helped, it didn’t solve all of the challenges of AI. Of course, there are limitations to AI models when dealing with multiple European languages, such as translation without context or differences in intonation. However, the team saw this as an opportunity to continue learning and innovating.
“My team never stops at one example when dealing with a new word or topic. Some European languages are more difficult than others,” Ross adds. “If you’ve ever been to Spain, you’ll know that Spanish is often spoken very quickly and we need to train AI very well to handle that.”
Expanding Galaxy AI required new collaborations across continents, but the work quickly expanded beyond European borders. SRPOL supported the Jordanian team’s efforts to teach a wide range of Arabic dialects using Galaxy AI, as well as the Brazilian team’s work on Latin American languages.
The importance of language nuances and cultural differences is on the radar of SRPOL product developers because they can all be noticed by the target: the end users.
“There are subtle differences between European cultures that affect whether something feels natural to the end user. For example, people in some countries expect to read prices in euros, while others are used to seeing them written in euros,” says Agata Maria Ruzicka, head of voice intelligence research at SRPOL. “If this cultural difference is not reflected in the translated text, the interface may feel less user-friendly. Implementing these ideas at a micro-level in interface design can make technology feel more natural in diverse cultures.”
“The team has been communicating and collaborating remotely in different countries for many years, which has led to the establishment of many effective communication channels,” says Marcin Mrugala, Head of Mobile Quality Assurance at SRPOL. “We were ready to do our part to enable Galaxy AI to reduce language barriers around the world.”
Technology to unite cultures
Managing and integrating diverse linguistic and cultural knowledge is a challenging task, but it is essential to Samsung’s vision for Galaxy AI: reducing the barriers that divide people based on language and culture, and enabling them to create deeper connections.
“We’re not just building the technology of the future, we’re building the hardware of the future,” says Mrugala. “Our best practices are designed to improve products based on differences between countries, but we fundamentally believe that our similarities far outweigh our differences and that our technology can bridge cultures.”
“Our goal is to bring people together, make their lives easier and simplify their daily tasks. We see our families using the voice recorder in new ways, and now we can call our friends in different countries and speak to them in their own language. It is magical to see this change in the world and to be a part of it. Galaxy AI has united the people of SRPOL and now we are uniting the world,” Rozica concluded.
“Beeraholic. Friend of animals everywhere. Evil web scholar. Zombie maven.”