Home › Academic publications › RAG time – Generative AI can track down and define many words – but not all of them

RAG time – Generative AI can track down and define many words – but not all of them

Michael FarrellPosted on Posted in Academic publications

Abstract

In this insightful article, published in the ITI Bulletin in July 2025, Michael Farrell explores the limitations of generative AI (GenAI) in handling obscure or technical terminology, which is a common challenge for professional translators. He proposes Retrieval-Augmented Generation (RAG) as a promising solution. RAG enhances GenAI by integrating external data sources, thereby improving factual accuracy. Farrell illustrates its potential through examples from translation workflows and highlights RAG’s role in empowering translators as informed decision-makers.

Published in

ITI Bulletin, July 2025; pp. 27-29.

Download

Download full article.
Alternative download.

LinkedIn0
About

Michael Farrell is primarily a freelance translator and transcreator. Over the years he has acquired experience in the cultural tourism field and in transcreating advertising copy and press releases, chiefly for the promotion of technology products. Besides this, he is also an untenured lecturer in post-editing, artificial intelligence, machine translation and computer tools for translators at the International IULM Unviversity, Milan, Italy, the developer of the terminology search tool IntelliWebSearch, a qualified member of the Italian Association of Translators and Interpreters (AITI), an Individual Member of the European Association for Machine Translation (EAMT) and a member of Mediterranean Editors and Translators.