Since 2024, Untranslate has been supporting one of the world’s largest automotive manufacturers, a true global leader in enhancing its in-car voice assistant technology. The collaboration reflects the evolving landscape of today’s automotive innovation. The scope and ambition of the project speak for themselves: multilingual voice command support across up to 38 languages and multiple regional variants.
Making Cars Speak Every Language
What is voice command support?
Voice command support enables drivers to control vehicle functions through natural speech, particularly in situations where hands-free interaction is essential. In practice, these interactions typically occur while navigating traffic, searching for nearby services, managing communication, accessing media, or operating vehicle features, all while maintaining focus on the road. Behind these simple interactions lies a sophisticated system trained to recognise real-life speech patterns, varied phrasing, and natural conversational flow. Our role extended far beyond translation. Our linguists interacted directly with the vehicle’s head unit, detached and installed in recording booths. They simulated real driving scenarios, tested alternative phrasing, monitored system responses, and most importantly, flagged inconsistencies. Based on these observations, their feedback was highly crucial and was integrated into backend refinements, improving recognition accuracy and overall user experience.
From translation to real interaction
Modern vehicles are evolving into intelligent, voice-enabled environments. Consequently, drivers expect to interact naturally with their cars while asking for directions, calling a contact, playing music, or adjusting vehicle features and receiving seamless responses.
To meet these expectations worldwide, the voice assistant software required training in up to 38 languages, including their regional variants. The project began with the translation and review of an extensive list of voice command prompts. However, the real complexity started after translation.
Beyond linguistic expertise, a smooth operational process ensured the project met the client’s needs. We recruited linguists who were native speakers of all required languages and dialects in Brussels. In many cases, multiple speakers per language were needed to also test variations in gender and accent. Moreover, several languages were tested on the same day, both on-site and remotely, requiring careful scheduling, structured workflows, and agile project management.
Why localisation is critical in voice technology
Voice interfaces require a higher level of linguistic sensitivity than traditional UI localisation. Spoken language is dynamic, informal, and highly context-dependent; in such contexts, literal translations are rarely sufficient. By working exclusively with native linguists and providing clear contextual guidance, we ensured that the voice commands reflected how users naturally speak in their languages and dialects. In particular, regional nuances, syntax variations, and cultural expectations were carefully considered during this process.
For a global automotive leader, this depth of localisation supports confident market launches and ensures that drivers worldwide experience intuitive, reliable voice interaction in their preferred language. At Untranslate, we helped our client achieve this goal by combining expert linguistic insight with hands-on AI training, rigorous quality checks, and seamless project management. The voice assistant now supports 38 languages and regional variants. It adapts to accents, natural speech, and user intent to provide a seamless experience, ensuring a truly global and human-centric experience.
Structured workflow, consistent quality
To ensure accuracy and consistency across all languages, we implemented a clearly defined workflow:
-
Translation and native review of all voice command prompts used to test the multilingual voice assistant
-
Terminology alignment across languages
-
Contextual guidance, including user intent and screenshots where required
-
Multiple days of on-site and remote training sessions with linguists different from the original translators
-
Continuous feedback loops with the client’s engineering teams
Close collaboration between project managers, engineers, reviewers, and on-site & remote linguists ensured smooth communication and efficient turnaround times. Our established multilingual network in Brussels and online enabled fast and smooth coordination, even when multiple languages were tested simultaneously.
Delivering global impact
After months of structured collaboration and testing, the multilingual voice assistant was integrated into the client’s vehicles in 2025.
The outcome:
-
High-quality multilingual voice support in up to 38 languages and variants
-
Enhanced readiness for global market launches
-
A smoother and safer driving experience
For drivers, this means the ability to interact naturally with their vehicle during navigation, communication, and vehicle operation in the language they know best and feel is most natural to them.
Driving the future of multilingual innovation at Untranslate
As vehicles become increasingly intelligent, language technology plays a defining role in user experience. This project demonstrates how linguistic expertise, technological collaboration, and structured project management combine to support innovation at scale. By bridging the gap between human communication and machine understanding, we ensure that every driver can interact naturally and confidently with their vehicle, regardless of language or region.
At Untranslate, we go beyond translating content — we help global automotive leaders build truly multilingual, voice-enabled mobility solutions. Our work ensures that speech technology is not only functional but intuitive, inclusive, and ready for the diverse markets of today and tomorrow.
Do you have a challenging translation or software localisation project? Our expert team is here to help!
Get in touch with us to learn more. Discover how we provide high-quality language solutions, including the linguistic quality evaluation of in-car multilingual support to our automotive clients.
