image

AI in the crossbase translation process

With the new AI features of Microsoft Azure and DeepL, translating with crossbase becomes even more effective.

Albrecht Bosler
CTO, crossbase mediasolution GmbH

Translating content into multiple languages is a key challenge for companies that want to sell products internationally. With crossbase, this process is fully optimized - from the precise determination of requirements to the integration of powerful machine translation services such as DeepL and Microsoft Translator. Thanks to innovative functions such as glossaries, crossbase significantly reduces the workload and ensures high-quality translations.

 

Standardized text and translation workflow

The translation process in crossbase determines the need for translations for all required foreign languages and manages the corresponding orders in a targeted manner. A sophisticated text workflow and the management of text modules make it possible to include only missing or outdated translations in the process. This significantly reduces the amount of text to be processed.

In addition, crossbase takes into account texts that are already being translated and reacts proactively to changes to the source text, for example by issuing warnings. In order to address specific requirements, a tool has been developed with which the respective contexts - e.g. for websites, e-catalogs, structures or print publications - can be precisely selected. Translations are only carried out where necessary.

Together with integrated translation services, crossbase offers a powerful solution that makes the entire translation process efficient, reduces the workload and ensures high translation quality.

Translation with DeepL

With the release of crossbase Release 5, DeepL was integrated as a translation service. DeepL, a self-learning online service for machine translations, delivers outstanding results thanks to its algorithms.

A wide variety of content such as product descriptions, product feature descriptions, marketing texts for websites or catalog texts - such as headings and chapter introductions - can be translated quickly and easily into the desired language. The translations are carried out directly in the familiar crossbase applications such as crossbase.pro, crossbase.trans or crossbase.web.

The automatic translation is created with just one click and can then be checked and confirmed. For extensive requirements, crossbase also offers the option of translating large volumes of text efficiently in a single step. Use requires a DeepL Pro subscription, which can be easily linked to crossbase. Existing subscriptions can also be integrated.

Integration of glossaries for DeepL

A common problem with machine translation is the lack of use of customer-specific dictionaries. This often leads to time-consuming correction loops, especially for special company terminology. To solve this problem, crossbase has integrated glossaries directly into the machine translation process.

This function allows glossaries to be defined with source and target languages. As a result, technical terms are translated correctly, which significantly reduces the correction effort and ensures consistent terminology.

Connection to Microsoft Translator

In addition to DeepL, crossbase also offers the option of using Microsoft Translator for cloud-based translations. This service expands the translation options with a larger selection of languages (see Microsoft Translator) and an integrated dictionary function.

The Azure-based Microsoft Translator is characterized by high security and scalability and complements the translation functions of crossbase with additional languages and features.

I look forward to a personal
consultation 
with you.

Herby Tessadri, Sales Manager

Call me +49 7031 9880-770
or write a message

Contact us

To prevent misuse of the form, we use "Friendly Captcha" to determine whether the user is a human being.

Thank you for your message! We appreciate your interest and will get in touch with you as soon as possible.

An error has occurred. Please try again.