TextShuttle offers both generic and customized (i.e. domain-specific) neural machine translation systems. Owing to TextShuttle's research proximity, the systems are subject to constant alignment with leading technologies. Apart from significantly increasing the productivity of the translators, the terminology integration option makes sure that your translations comply with internal policies and the post-editing overhead is kept to a minimum.
TextShuttle enables us to run domain-specific NMT on our local servers, effectively integrating the technology in the Across workflow.
Beatriz Gonzalez, Senior Business Analyst at Migros Bank AG
TextShuttle, a spin-off of the University of Zurich, offers the creation and maintenance of neural translation systems on the basis of state-of-the-art language technology. The systems can be configured easily. The models and parameters can be adapted to customer needs during operation, and updates can be installed in the system without interrupting the operation.
On request, the machine translation systems can be set up in the customer's on-premise network. In this way, the texts are not translated on third-party servers at an unknown overseas location, but locally in the customer's own IT infrastructure. Thus, customers have full control over the access to the translation systems and the data processing. This guarantees maximum data protection, data security, and confidentiality.
The TextShuttle systems make use of the latest findings from the field of technology research and enable the integration of corporate terminology.
Our domain-specific translation systems are created in close collaboration with the customer. The customer's needs and preferences are taken into consideration, as are special language variants, styles, and any required special characters. Customer can directly integrate their terminology to ensure full control over corporate translation specifics. Translation memories are integrated during the system setup and in the context of re-training measures.