Glossary

MT engine

An MT engine (Machine Translation engine) is a software system that automatically translates text from one language to another using computational models. It forms the core of machine translation (MT) technology and can be based on rule-based, statistical or neural network approaches.

Description

Modern MT engines are built using neural machine translation (NMT) architecture, which applies deep learning to understand linguistic structure and context. By processing large volumes of bilingual text, these engines learn how to produce fluent and accurate translations that mirror human language.

Each MT engine can be customized for specific domains, languages or clients – for example, adapting to the terminology and tone used in legal, life sciences or marketing content. This specialization improves translation quality and reduces post-editing effort for linguists. At RWS, Language Weaver is the enterprise-grade MT engine that combines neural models, adaptive learning and secure data handling. It integrates seamlessly with Trados, TMS and API-based workflows, allowing organizations to translate high volumes of content quickly while maintaining data security and linguistic integrity.

Related terms

AI translation Language Weaver MT optimization Machine translation (MT) Neural machine translation (NMT) Post-Editing Trados Team

Transform how the world understands you.

Discover the technology and expertise that powers the world’s most trusted brands.
See what's possible