Translation technology is quickly changing with new tools released each day. Being vendor agnostic allows us to integrate with our client's requirements rather than a tool provider we are aligned with. To provide such objective guidance requires a company that is familiar with all of the technology in the localization industry and is able to maintain a lab environment to constantly evaluate different platforms. Our engineers are familiar with all major industry tools, supporting all formats and eager to help with any questions you might have. Contact us today with any of your translation tools questions at email@example.com
Neural Machine Translation
Neural Machine Translation (NMT) is a breakthrough approach to machine translation that uses a large artificial neural network to predict the likelihood of a sequence of words. The cost for translating documents using Neural Machine Translation can be up to 50% less than when using a traditional TEP (translation, editing & proofreading) process. Files are first “pre-translated” via an enterprise-grade NMT engine and previously translated content is located throughout the document. A professional linguist then edits the translated text looking for obvious errors with terminology, grammar, and sentence structure. Neural Machine Translation is not suitable for all languages or materials. Regulated materials, sensitive data, marketing collateral and other customer-facing content should always be translated and proofread by professional human translators. To learn more, contact us today at firstname.lastname@example.org
Translation Memory (TM) is a translation database technology that stores and reuses previously translated segments in order to aid human translators. Using Translation Memory can dramatically reduce the scope, cost and timeframe needed to complete translation and localization projects. Because previous workings are suggested to linguists, translations are also much more consistent than with projects not using translation memory. TranslateNow supports all major TM platforms and always provides updated TMs to our clients after each delivery.
OCR: Optical Character Recognition
Optical Character Recognition (OCR) is a tool commonly used in the translation process. OCR is used to convert files such as PDF, images and scanned content into an editable format like Microsoft Word. Our powerful OCR technology supports 75 languages and we even have the ability to work with multilingual handwritten content. Our professional desktop publishing team recreates the exact format of your translated files ensuring the highest quality output.
Case Study: 2.5m Words in 45 Days
TranslateNow recently received 10,000 pages of paper log books from a major airline manufacturer that needed to be translated in order to complete the sale of a very expensive aircraft. Unfortunately, all of the logs were unorganized and a mix of Portuguese, French, German and English. The team estimated 2.5 million words needed to be translated at an initial cost of over $750,000. Additionally, we anticipated needing at least six months to complete the project. TranslateNow’s engineers suggested scanning and OCRing the logs and then processing them through NMT followed by an edit by several teams of professional linguists. The results were astounding. By using NMT we were able to reduce the workload by over 70% and our delivery time by five months meeting the client’s aggressive deadline and budget.