BNP Paribas, “the bank for a changing world”, is revolutionizing its customer experience offering by introducing the unique, fit-for-purpose Voice for its digital assistant, Telmi. Telmi’s voice has been tailored to reflect and carry BNP Paribas’s brand identity across channels. As such, Telmi’s voice is entirely ubiquitous and available across several touchpoints: omnichannel contact centre, in-app solutions, and smart speakers such as Google Home.
Not stopping there, BNP Paribas has also created the unique and personalised Voice of HelloïZ, the counterpart of Telmi for its HelloBank! brand. True, next-level brand differentiation.
MVP is a cloud-based IVR, built on top of Genesys technology and natively integrated with Orange Contact Centres. A highly customisable solution, MVP informs and advises customers swiftly and efficiently, while also facilitating broader interactive features with natural language and text-to-speech capabilities. Voxygen’s Expressive Speech solutions are integrated into MVP in seven different languages. Voxygen’s Studio provides user-friendly voice tuning features.
With two billion passengers transiting every year, SNCF Gares & Connexion needs to integrate Traveller Information throughout the customer’s whole journey: railways stations, interactive kiosks and ticket machines, Interactive Voice Response (IVR), mobiles phones, etc. With up to 1200 simultaneous announcements made by Voxygen’s “e-Mone” (the digital twin of Simone Herault, SNCF’s Voice) in six languages, SNCF needs to automate and enrich Traveller Information throughout 1500 French train stations. The EVA Project fits within the strategic vision of providing richer Traveller Information, consistent and homogeneous across channels, in stations or in transit.
For several years now, EDF has defined its own Brand Voice and communicates across many media channels with a voice talent who really personifies their brand identity. With the rise of voice-based interactions, EDF was willing to shift from static messages using pre-recorded speech to richer and more personalised interaction with dynamic content generated by Text-to-Speech technology. To facilitate this shift, all while preserving the existing EDF vocal identity, Voxygen created EDF’s digital Brand Voice. This unique voice is now operating in several voicebots and IVR systems, ensuring a unified and homogeneous voice experience for customers across all of EDF’s contact centres.
For the launch of Season 3 of Versailles, a successful Canal + TV series, BETC decided to recreate the voice of King Louis XIV of France, more than 300 years after his death. Thanks to data collected on his known physical characteristics and from his personal health journal, Voxygen synthetic speech experts were able to develop a voice no one had ever heard before. BETC staged an #audiencewiththeking on social media – an opportunity for the public to put their questions to the king himself and even to hear his replies. The Louis XIV voice has also been used to welcome visitors to Virtually Versailles, a digital experience offering a tour of the Palace of Versailles itself. Vive le Roi ! Vive la voix !
Météo France develops innovative weather forecast services available across many channels such as IVR, voicebots and mobile apps. Since voice remains the simplest interface to convey such information, Météo France contracted Voxygen to create a voice in order to deliver accurate weather information in a natural and fluid manner. To address a wide variety of audiences – from individuals needing general information to professionals with very specific needs – Voxygen designed a personalised voice capable of handling a very rich vocabulary, including terms specific to maritime, mountain or agricultural weather forecasts. The voice is currently deployed in several services, enabling Meteo France to deliver high-value and continuously updated weather information.
AlloCovid is a toll-free automatic phone service that aims at screening and following-up on those who have been exposed to the Covid-19 virus and who have been considered to be at very high risk. AlloCovid contributes to the real-time detection of patients that might go on to develop a severe form of the disease due to their pre-existing conditions. AlloCovid has developed AI-based virtual phone assistants that operate 24/7 and can respond to more than a thousand people at the same time. Voxygen is supporting this project, led by Inserm, Allomedia, eVoyageurs SNCF and Privacy Impact, by making available its Voxygen Cloud service with expressive voice technology for digital phone assistants.
Voxygen has developed a new methodology which allows for fast language development in the context of Speech-to-Speech translation. The development of a Speech-to-Speech solution incorporates three main components: speech recognition, translation and speech synthesis. The MAGMAT technical framework relies on utilising common resources (linguistic knowledge, phonetic alphabets, text and acoustic data) for the three technological components and is based on AI and machine learning techniques.
Languages: Turkish, Swahili
Partners: Airbus, LIUM (Laboratoire d’Informatique de l’Université du Maine).
