Microtalks
Track: Microtalk | M1 | Intermediate | MICRO TALK |
Wednesday, June 4, 2025, 9:00am – 9:30am
Held in: Live 3
Presenters:
Henrik Kühnemann - Translator Scandinavia AB
Andreas Ljungström - XTM International
Chiara Scaldaferro - GoStudent
Nicolas Underwood - Lionbridge
Host: Daniel Goldschmidt
In this session, we will offer several quick and dynamic presentations covering topics and ideas too interesting to ignore!
From Numbers to Narrative: Structuring Data to Showcase ROI Impact presented by Andreas Ljungström, XTM International
In this presentation, we will explore the vital connection between decision-making and meaningful information which we call data structures, emphasizing the importance of demonstrating value in all aspects of business. You will learn how to make informed choices regarding technology, team dynamics, and partnerships while navigating the pressures of maintaining business identity. By establishing a solid data foundation, organizations can articulate the impact of their decisions and demonstrate their value effectively. Regardless of your role or industry, the insights shared will provide actionable strategies for defining and showcasing your value in a rapidly evolving landscape. Join us to unlock the potential of purposeful decision-making.
Takeaways: Decision-Making Framework: a structured way to make decisions using data to evaluate technology investments, team setups, and partnerships for better choices. Data Foundation: how to build a strong data infrastructure that shows the impact of business decisions, improving communication of value to stakeholders. Value Communication Techniques: practical tips for clearly expressing their value in a changing business environment, helping them stand out and adapt confidently.
No Strings Attached: Towards a Standard of Data Extraction for Document Translation presented by Henrik Kühnemann, Translator Scandinavia AB
There is a lot of talk about long context translation and translating documents rather than segments. With the advent of RAG new perspectives on the document emerge. Where the RAG perspective is only concerned with exposing the content to the LLM, our task doesn’t stop there. We need to ensure a formally loss-less roundtrip. This MicroTalk will focus on the need of turning the TMS upside down, establishing a new taxonomy for content types and outline challenges and possibilities with a multilingual exchange format that extends rather than reduces.
Takeaways: Attendees will take away a better understanding of digital content types.
The Rise of the Humans: Time for Translators to Take Control of the Machines presented by Kasia Druszcz, Lionbridge Games
Initial uses of Large Language Models (LLMs) in translation have aimed at enhancing pre-translation processes to reduce translator effort while maintaining quality. This method is most effective with homogenous content that can utilize predefined prompts. However, game content is highly varied, presenting unique challenges across different languages. Despite this, targeted inputs and effective prompting strategies can significantly aid the translation process. The key is to empower translators, who possess deep content knowledge and linguistic expertise, to leverage LLMs without needing to become expert prompt engineers. This presentation will explore future localization strategies that balance human creativity with LLMs efficiencies.
Takeaways: Gain insights into how LLMs can be effectively integrated into the translation process; learn strategies to empower translators with LLMs, enhancing both efficiency and quality; discover where the future of game localization and language is heading and how to stay ahead in this evolving field
One Person, Global Reach: Scaling Localization with Automation for Industry Newcomers presented by Chiara Scaldaferro, GoStudent
Takeaways: The secrets to building an efficient localization strategy as a one-person team; how to control quality across multiple languages using a trusted Language Lead system; exploring integrations and AI to streamline workflows, reduce manual tasks, and ensure scalability.
