Machine translation solutions

Adaptive AI-powered translation solutions for fully automated localisation.

Advanced solutions for global enterprises & organisations

Extensive research in Natural Language Processing, technological advancements and computational linguistics training have resulted in the outstanding development of sophisticated translation technology, producing grammatically coherent translated output and showing promising results in a variety of language pairs.

To manage increasing translation volumes and the need for fast turnaround times, we now support a cutting-edge high-quality machine translation solution using Neural Machine Translation (NMT) technology. Thanks to continuous progress in quality and output and seamless integration with our translation environment, Machine Translation is the ideal tool for high-volume, organised, structured content in certain language pairs and documents where timelines and budgets are a concern.

Our Machine Translation solutions

We have a number of innovative technology solutions that make localisation seamless and cost-effective

Fully automated Machine Translation

By removing human intervention in the translation process, delivery times are drastically faster than human-only translation. Our automated neural machine translation tasks are executed within our Translation environment to ensure safe and seamless integration with any application or workflow and enable a streamlined process and consistent translation outcomes across your organisation.

Our data centre is hosted in Australia and complies with strict security and privacy requirements. The Neural Machine Translated (NMT) output occurs through our cybersecurity secure server. The NMT Engine is customised for each client. It applies first the client-specific terminology stored and curated in our Term Base and Translation Memory for that particular client before prepopulating the remainder with NMT output.

This service is the most cost-effective way of translating large volumes of text in short timeframes and our customers can have access to this secure NMT system at any time through our secure Customer Portal, pass their content through the engines themselves and receive automatically the translated content.

Learn More

Customised Machine Translation

We can develop and train a private neural machine translation engine for you provided that you have enough translated training data. By training your custom NMT engine with your translated data, the neuronal model learns and understands your company’s language using Artificial Intelligence and Deep Learning. The output in translations will better reflect your company’s language and style guide than a non-customised NMT engine (read more about the difference).

This solution is relevant for companies that have already large amounts of quality translated data and that need regular and large volumes of translations. One custom NMT engine can perform translations in one language combination and direction, e.g. one engine for English into Spanish (LA). The engine can be re-trained if there is new data to further improve its performances. We advise gathering at least 100,000 segments pairs of translated data before training an engine as a rule of thumb. Best results are observed for particular projects depending on the size, the quality and the relevance of the training data.

We can provide our custom NMT service in a fully automated workflow where you submit your content through our secure customer portal and you receive the translation in the same place within minutes.

Learn More

What is machine translation

It is the translation of content without human involvement. Pioneered in the 1950's machine translation has significantly evolved to include new models that include cutting-edge technologies such as artificial intelligence and machine learning. There are 3 types of machine translation as well as hybrid systems, a mix of them.

The rule-based system functions with a combination of language and grammar rules plus dictionaries for common words. Specific glossaries or term bases are created to focus on certain industries or disciplines or for a certain client. Rules-based systems typically deliver consistent translations with consistent terminology when trained with a specific termbase.

Statistical systems do not have implemented language rules. Instead they "learn" to translate by analysing large amounts of data for each language pair. They can be trained for specific industries or disciplines using additional terminology relevant to the sector needed. Typically statistical systems deliver more fluent-sounding but less consistent translations.

Neural Machine Translation (NMT) is a new approach that makes machines learn to translate through one large neural network. The approach has become increasingly popular amongst researchers and developers, as trained NMT systems have begun to show better translation quality in many language pairs compared to the statistical approach.