XMAT® Machine Translation Platform Launches New Feature Utilizing RAG Technology to Reference Language Asset Data
The machine translation platform "XMAT" has added a new feature that utilizes RAG technology to reference language asset data. This enhances the accuracy and consistency of generative AI translations, allowing for efficient use of glossaries and past parallel data. This update significantly contributes to streamlining translation work and stabilizing quality.
📋 Article Processing Timeline
- 📰 Published: April 1, 2026 at 20:00
- 🔍 Collected: April 1, 2026 at 16:47
- 🤖 AI Analyzed: April 21, 2026 at 16:47 (479h 59m after Collected)
■XMAT Update Details
Product details page: https://ldxlab.io/xmat
Added Feature:
Implemented a function to reference language asset data when performing generative AI processing in the document translation feature.
*Language asset data: Refers to glossaries and parallel data used in translation.
In XMAT's document translation, in addition to generative AI translation and grammar correction, users can also specify and execute their own prompts. While this prompt function allows for detailed specification of terminology and writing style during translation, it also involves a significant amount of effort to list all terms and expressions to be unified and reflect them in the prompt. RAG (Retrieval-Augmented Generation) is the technology that solves these challenges.
XMAT now utilizes RAG technology, making registered language assets available as reference information for AI processing. Just like selecting a translation engine or language, users can simply select the necessary language assets to reflect their content in the output. By combining this new feature with the conventional document translation function, more highly accurate and consistent translations can be achieved.
■What you can do with Document Translation × RAG Technology
1. Unify expressions using past parallel data
By referencing expressions from previously translated documents, the same phrasing and translated terms can be prioritized. This prevents inconsistencies in terminology and contributes to stabilizing quality.
2. Accurate translation reflecting glossaries
Product names, technical terms, and company-specific expressions are automatically referenced from glossaries, and specified translated terms are accurately reflected. Significant reductions in revision work and review man-hours can be expected.
XMAT will continue to evolve by actively incorporating the latest technologies and promptly reflecting user requests into features, thereby contributing to business efficiency.
■About XMAT
Details page: https://ldxlab.io/xmat
Main features:
● Cross-platform use of multiple AI/translation engines with "Quick MT Text Translation"
Various engines such as Minna no Jido Honyaku, Claude, Gemini, and GPT can be selected according to purpose.
It also supports generative AI processing such as summarization, style conversion, and paraphrasing in addition to translation.
●Excel/PowerPoint/PD