From Glossaries to AI Pipelines: Mine, Structure & Connect Terminology in Intelligent L10n
AI Use Cases SIG Meeting - August 2026
Terminology has always been the backbone of consistent, high-quality localization, but AI is fundamentally changing how we build, manage, and deploy it.
In this session, three experts take you on a journey from the foundations of glossary extraction to the cutting edge of AI-powered terminology ecosystems.
They will discuss how hybrid AI approaches outperform single-method extraction, how a future-ready termbase is designed with LLMs and MT in mind from day one, and how open protocols like MCP are closing the gap between linguistic expertise and intelligent automation.
If you're modernizing your terminology program or exploring AI integration, this session offers a rare end-to-end view of terminology in intelligent localization.
Takeaways
- No single AI method is enough: combining multiple extraction engines produces more reliable, production-ready glossaries than relying on any one approach alone
- Termbase design is a strategic decision: metadata structure and contributor workflows directly determine how well your terminology performs in LLM and MT environments
- TAG & MCP makes AI-termbase integration scalable: open protocols allow AI agents to access and leverage terminology knowledge without complex custom engineering
- Terminology programs are living systems: from extraction to structure to AI connectivity, each layer compounds the value of the next
Who should attend
This session welcomes anyone involved in localization quality, content, or language technology, whether you're a seasoned terminology expert or just beginning to explore its impact:
- Localization managers who recognize the importance of consistent language across markets but aren't sure where to start with terminology
- LQA specialists evaluating AI tools for glossary and terminology workflows
- Globalization leads looking to modernize, scale, or future-proof their termbase programs — and exploring how LLMs and MT can be better grounded in controlled terminology
- Localization managers struggling with quality. If inconsistent terminology is affecting your translations and you're looking for proven, practical approaches, this session is for you
- Terminologists who want to understand how their work fits into a modern, AI-powered localization workflow
If you believe terminology is the foundation of localization quality and are ready to take it further (from first steps to AI integration), this session will meet you where you are.

Balázs Benedek is CEO of Easyling, an AI-assisted translation and localization platform covering websites and documents using proprietary Translation Proxy technology. He works closely on product and engineering, with a focus on applying AI-powered automation to real-world localization workflows.

Rainbow Liao is Localization Project Manager at Zoetis. Rainbow is a localization expert at Zoetis, where she launched and leads the company’s terminology program. Passionate about all things terminology, she focuses on harnessing AI to unlock the value of structured language resources for multilingual content in animal healthcare. Rainbow’s work bridges technology and linguistics, helping Zoetis deliver clear, consistent communication across global markets.

Christian Lang combines deep language expertise with strong technical know-how, shaped by his studies in Japanese Studies and Translation as well as his cross-disciplinary research on deep learning approaches to term and relation extraction. He first discovered terminology management while working as a freelance translator for the European Patent Office. Since then, he has been exploring the intersection of language and technology, with a focus on machine translation, automatic term extraction, knowledge graphs, NLP in general and deep learning/AI specifically. Since 2022, he has been part of Kaleidoscope’s Professional Services team, where he also founded and leads Kaleidoscope AI Labs.
GALA SIGs are global communities that connect members on topics of shared interest led by volunteer moderators. Meet the moderators for the AI Use Cases SIG.
Dr. Belén Agulló García, Terra

Belén is a multimedia localization professional with over a decade of experience in the localization industry in various roles in multiple companies, mainly language service providers, research and consultancy firms, and technology service providers. In her professional career, she has touched upon almost every role in the industry, including linguist, project manager, quality program manager, language technology deployment manager, researcher, educator, customer success manager, and vendor relationship manager and community builder, among others. This unique mix of experiences gave her a unique and deep understanding of the industry. In recent positions, she worked as AI Strategist and Innovator. Her focus has been to lead teams in devising and implementing long-term AI strategies, foster strategic partnerships, and define go-to-market strategies for AI services across key verticals. Her current role is Executive Consultant, Innovation at Terra, where she keeps bringing her passion for localization, processes, and language technology to customers around the world. With an academic foundation rooted in Translation and Interpreting Studies from the Universitat d'Alacant, an MA in Audiovisual Translation, and a Ph.D. in Language Interpretation and Translation from Universitat Autònoma de Barcelona, she has not only pursued knowledge but has also contributed to her field through a user-centered study on accessible subtitles in virtual reality content.
Marcus Danei, SAP

At SAP, Marcus Danei is responsible for defining and implementing the strategic direction for SAP's language technology. Marcus has over 24 years’ experience in the software and localization industries, and has worked in a range of areas within SAP, from introducing Arabic as a standard language, through translation production management, defining and maintaining standards for release management and technology, to translation business partner management.
SAP’s strategy is to help every business run as an intelligent enterprise. As a market leader in enterprise application software, we help companies of all sizes and in all industries run at their best: SAP customers generate 87% of total global commerce. Our machine learning, Internet of Things (IoT), and advanced analytics technologies help turn customers’ businesses into intelligent enterprises. SAP helps give people and organizations deep business insight and fosters collaboration that helps them stay ahead of their competition. We simplify technology for companies so they can consume our software the way they want – without disruption. Our end-to-end suite of applications and services enables business and public customers across 25 industries globally to operate profitably, adapt continuously, and make a difference. With a global network of customers, partners, employees, and thought leaders, SAP helps the world run better and improve people’s lives. For more information, visit www.sap.com.
Xavier Maza-Cid, iDISC Information Technologies

Xavier Maza is Chief Client Solutions Officer at iDISC, where he leads sales, customer service, and quoting operations for global language and technology projects. With over 30 years in translation and localization, he combines strategic vision with hands-on industry expertise. Xavier is focused on integrating AI-driven solutions into language services, bridging human talent and technology to deliver smarter, scalable, and secure communication. Passionate about team leadership and innovation, he closely follows developments in AI, machine translation, and cybersecurity. He speaks Catalan, Spanish, and English fluently, with working knowledge of Danish and Russian, and is an avid scuba diver.
Zalan Meggyesi, Easyling

Diplomat by training, solutions engineer by trade, and usually the person brought in when ideas need to become something real. I work at the intersection of language, systems, and automation, with a strong focus on rapid prototyping. Recently, that’s meant hands-on experimentation with AI as a prototyping accelerator: turning rough ideas into specs, specs into prototypes, and prototypes into decisions. In this SIG, I frame the discussion, surface practical questions, and help the group extract useful patterns from real experience, so we leave with clearer signals on how to leverage AI in our favor.
Ziyi (Ian) Zhang, Zoetis

Technical Project Manager at Zoetis with expertise in machine translation evaluation, AI training and localization workflow automation. Skilled in designing and applying evaluation metrics tailored for the medical and pharmaceutical domain, with a strong focus on accuracy and terminology. Experienced in building AI-driven solutions, integrating terminology management tools with AI translation. Passionate about driving innovation at the intersection of language technology and global communication.
Join GALA SIGs
Our special interest groups are global communities that connect members on topics of shared interest. Explore and join GALA SIGs.
For More Information: