Sometimes you might not need a perfect translation. You may just want to get the gist or translate some bulky documentation in a short time on a limited budget. This is where AI translation services come into play, with human translator post-editing to fine-tune the final output. Our automated process makes it quick, easy, and painless to get your translated copy.

How does an online translator work?

Dialogue’s Translation Management System platform comes bundled with 60+ machine translation engines to pick from, including DeepL and Microsoft Translate, each with different specialisms and strengths. Our Autoselect function scans your documentation and chooses the engine most suited to your subject matter and language combination. You can also attach a term base or translation memory for a more personalised final output.

Professional Advice

If you’re unsure whether machine translation is right for you, our expert linguists will guide you.

More than just Google

Our TMS has over 60 engines to choose from and an Autoselect function to help you decide on the best one.

Personalised Approach

If you already have preferred translations, our human linguists can work these in for greater personalisation.

Dialogue's Machine Translation Process

Step 1.

Upload source content to Translation Management System's Machine Translation in editable format. The autoselect function selects an appropriate Machine Translation engine.

Step 2.

Integrate any relevant translation memory to the job for improved output. Content is processed by the engine and auto-translated in an instant.

Step 3.

A professional human translator reviews the Machine Translation and edits where required for a more polished final version.

Step 4.

Translated content uploaded to Translation Management System in original format for you to review.

Step 5.

Feedback to ensure improvements and maintain quality for future projects.

Machine Translation FAQs

Machine translation, or MT, is translation completed by computer software built from a large database of existing translations and continual input. There is an array of MT engines in the market, mainly built around neural translation technology which enables the software to analyse and translate text in context for greater accuracy. Statistical MT also forms a part of some engines, translating more on a word-for-word basis, with its strength in technical text. Click here for our webinar on machine translation. 

Machine translation technology has improved immensely over the past few years, thanks mainly to advancements in neural technology. Having said this, while good for gist and many types of content, if you are looking for a 100% accurate & reliable target text, using raw machine translation is not often recommended. Post-editing by a professional linguist will improve the final flow and accuracy.

Machine translation works well for some content (e.g. for gist, internal communications and some technical texts), and not so well for other material (e.g. marketing and creative). Also, some language combinations are stronger than others. Feel free to send us the content you’d like translating and we can advise you as to whether machine translation is a suitable option or not.

We source machine translations from multiple different machine translation engines (e.g. epL, Systran, Google Translate, Microsoft Translate). We use our translation platform’s inbuilt AI to analyse the source text and select the best translation engine based on the source domain and post-editing data collated from all over the world. This allows us to get the most accurate machine translations for your project.

Machine translation costs vary depending on how rigorous you would like the post-editing to be, and the language combination used. A ballpark figure for languages like French, Spanish and Italian would be between £40 per 1,000 words for raw machine translation and £60 for machine translation with a high level of post-editing.

We use a combination of 10+ machine translation engines to provide the most accurate translations possible. Included in this is DeepL, Systran, Google Translate, Microsoft Translate, and more. These engines can be customised with specific terminology and termbases to adapt the translations to your preferences.