Thanks to the deep integration of ReadSpeaker technology, CBTL’s EVOLUTION³ production platform has been setting the standard for the creation of audio-enhanced and multilingual learning content for years. Over the years, the requirements of many demanding clients have been incorporated into the partnership and the platform. This has made it even easier to generate high-quality audio content and has enabled more efficient management of multilingual content directly within the authoring tool. There is no need for complex external data processing or management. At the same time, maximum data security is guaranteed. International clients such as CLAAS, who require complex, technically precise and globally scalable training content, benefit particularly from this.

Precision and brand identity in many languages
Global companies such as CLAAS, one of the world’s leading manufacturers of agricultural machinery, are constantly faced with the challenge of producing and managing multilingual content. This requires ensuring not only technically accurate but also brand-compliant pronunciation of product names, abbreviations and technical terms across more than 20 voices. However, traditional text-to-speech solutions often fail to reproduce and pronounce brand names and technical terms correctly and consistently. This poses risks to comprehensibility, credibility and subsequent learning success.

“Thanks to the support of native speakers and country teams, we were even able to successfully replicate non-Latin languages such as Ukrainian. We overcame this challenge through creative solutions, combining ReadSpeaker technology with human expertise,” explains Matthias Ruch, Training Technology Engineer at CLAAS.
For global companies with a broad portfolio of products and services, the manual creation, adaptation and maintenance of content has proven to be time-consuming and inconsistent. By combining EVOLUTION³ and ReadSpeaker, a solution has been created for rapidly changing situations. In this setup, an author can override the default pronunciation from the dictionary without having to make manual corrections. In collaboration with ReadSpeaker’s linguists, CLAAS has created a master dictionary. This functionality ensures consistent pronunciation and guidelines for defined terms for all authors. This preserves brand identity, supports linguistic complexity and ensures seamless integration into the existing production environment.
Customised language technology integrated directly into EVOLUTION³
CBTL and ReadSpeaker have jointly developed a bespoke solution tailored to CLAAS’s specific requirements. A custom-trained pronunciation dictionary containing thousands of rules ensures that terms are pronounced accurately and consistently in the respective languages. By integrating ReadSpeaker’s speechCloud API into the EVOLUTION³ production platform, authors can create audio content directly within their familiar working environment without interrupting their workflow. For customers with particularly high security requirements, both the ReadSpeaker solution and that of CBTL can be hosted and operated within their own IT infrastructure.
“The support provided by ReadSpeaker’s linguistic team was crucial in achieving excellent results – without the lengthy trial-and-error process of testing letter combinations that is currently common with many TTS solutions,” emphasises Matthias Ruch, Training Technology Engineer at CLAAS.

Key technical aspects of the integration:
- Seamless API integration: Text is sent in real time from EVOLUTION³ to the ReadSpeaker service and returned as audio in the desired format.
- Customised pronunciation dictionaries: A dataset containing specific rules for each language and voice, in which the correct and consistent pronunciations of brand and product names are defined.
- Secure data processing: There is no unauthorised data leakage to external services, ensuring that data integrity is fully guaranteed on this secure platform.
Authors benefit from maximum efficiency, quality and flexibility.
With EVOLUTION³, you can create audio-enhanced, multilingual learning content more quickly, cost-effectively, securely and flexibly than ever before.

By using ReadSpeaker, you can generate professional audio content in over 50 languages. The integrated user interface ensures consistent voice and pronunciation quality without any interruptions or loss of quality.
Fast and cost-effective audio production
Computer-generated voices are available 24/7.
Voice-overs are produced without any waiting times or dependencies.
Consistent technical and vocal quality
Always speak in the same professional tone, regardless of how you feel that day.
No cutting out of breathing sounds or other artefacts.
Multilingualism and accessibility
Over 50 languages and 200 voices are available.
Automatically generated subtitles support accessible learning.
Flexibility and timeliness
Content can be quickly adapted, translated and dubbed into several languages.
Localisation is accelerated, and maintenance and updating are simplified.
Pedagogical freedom and corporate identity
Learning materials are created using precise corporate design templates.
A wide range of teaching methods for long-term learning success.



