[LIVE WEBINAR] From TMS to AI: The Next Evolution in Translation Workflows

"AI dominates the narrative across almost all industries, and doubly so in our segment, but we are here to stress that translators are not being replaced, but empowered. To us, AI mirrors the transformative shift experienced during the introduction of Translation Management Systems in the 1990s: AI offers new tools that free human experts from repetitive labor, letting them contribute more meaningfully. Our presentation explores how AI and LLMs (Large Language Models) are transforming translation workflows. We aim to share how practical implementations of AI can drive efficiency, improve translation quality, enhance client satisfaction, and empower linguists to do more faster."

In this webinar Zalan Meggyesi and Peter Argondizzo explore client lifecycles across three phases: pre-translation, mid-translation, and post-translation. In the pre-translation phase, a frequency analysis model supported by LLMs identifies glossary terms automatically during content extraction. This approach ensures that custom, impactful terminology guides machine translation (MT) from the outset. The mid-translation phase applies automatic MT for a quick first pass, followed by AI-powered quality evaluation to prioritize sections that require human review. LLM-generated suggestions aid in enhancing flagged segments, allowing for focused human intervention that maintains original voice and style. In the post-translation phase, low-quality machine translations are refined by human translators, using LLM recommendations, and the revised content is fed back into the MT engine to continuously improve performance.

The key takeaway is that AI enhances human expertise by reducing repetitive tasks, streamlining proofreading and review, and allowing translators to focus on nuanced linguistic challenges.

Zalan and Peter will discuss the impact of these advancements on client ROI, highlighting how AI enables expansion into new translation initiatives, benefiting both translation agencies and clients alike. During their presentation, they will showcase examples of LQA (Linguistic Quality Assessment) reports, demonstrating improvements in both quality and time-to-market, as well as share our insights into the selection of LQA models and their best integration into existing workflows.

Host organization: Globalization and Localization Association