Conference Program 2024

memoQ co-CEO, Peter Reynolds, will give a progress report on memoQ’s activities over the last year.

Neural machine translation (NMT), even with the amazing jump in quality that it brought over older MT methods, is being completely outshined by the promise of generative AI and Large Language Models. Without a doubt, these powerful new technologies are having a dramatic impact on the work of translators, interpreters, and everyone in localization. Sure, these incredibly flexible systems are limited by the amount and quality of available training data, but they’re also limited by our imaginations: after all, they don’t do anything until they get a prompt from us.

In this session, you’ll get the basics on what LLMs are and how they differ from NMT, up-to-the-minute information about how far they’ve come in the past year, and how they are being used in loc now, how they will continue to develop and further change the language services industry, and how you can choose to react to these changes.

Today, there are new AI-powered tools and solutions popping up every day. But is the sky really the limit when it comes to their potential to transform translation and localization?

This presentation offers a holistic view of AI’s impact on translation and localization within the context of the multilingual content lifecycle. Focusing on practical use-cases, we provide insights into how AI is leveraged in fields such as terminology management, translation, machine translation, automatic post-editing, quality assurance and estimation, etc.

The session commences with a quick market overview, outlining a burgeoning landscape shaped by cutting-edge technologies.

• But how can these technologies be integrated into existing processes, and how do they change these processes?
• What is the impact on business cases and the role of stakeholders
• What does this mean for our business models?
• And does AI really hold its promises when it comes to quality?

Starting from traditional processes, we outline where and how AI technologies can be integrated, what we can gain from this and what the requirements and risks are. We further provide insights on what this means for both translation service providers and enterprise language departments, addressing the central question they are faced with today, namely, how they can stay relevant in the future.

The presentation concludes with a forward-looking perspective on how embracing AI can empower language professionals to adapt, evolve, and stay indispensable in an ever-changing and increasingly technology-centric landscape, and what the prerequisites are for unlocking its full potential.

Yamagata Europe has dedicated over two years to designing and implementing a tailored translation process for the marketing communication of our valued customer, Mazda Motor Europe.

With a strong focus on integrating different content repositories, centralizing content streams, and automating manual and repetitive tasks as much as possible, Yamagata aims to improve efficiencies for every stakeholder in the supply chain and maximize local market satisfaction.

As the central TMS, memoQ has proven to be a key success driver in this transformation process.

To increase time-to-market and minimize the risk of human errors, Yamagata has leveraged memoQ’s APIs to automate data exchange between Mazda’s marketing automation platform, Oracle Eloqua, and web CMS Optimizely (formerly Episerver) on the one hand, and memoQ on the other.

Secondly, using memoQ’s newly released in-country review tool, Yamagata manages to intensify cooperation with Mazda’s local content owners.

Stakeholders from 22 national sales companies were onboarded and supported extensively in selecting and instructing dedicated linguists representing the local markets.

Receiving their feedback in the in-country review tool enables Yamagata to maintain translation memories in accordance with their needs and provide continuous training to the involved linguists.

Finally, Yamagata relies on memoQ’s analytics and reporting capabilities to collect relevant ROI data and share valuable insights with Mazda about their marketing localization program.

In this presentation, we will share insights on the challenges faced and the lessons learned throughout our journey.

See how UNICORN has evolved since last year with new integrations, interpreting enhancements, and a sandbox environment for early adopters.

When AI became commonly available, many suggested that it would be suitable to do Term Extraction, a task that otherwise would take a human translator or terminologist a substantial amount of time.

The question that had to be asked is not whether it can do it, but whether the AI can do it as well as a human. To answer this, we set up an experiment, which we believe to be the first of its kind in several ways, asking five translators to extract terms from a corpus of software strings, and then asked eight different AIs to do the same, turning the general guidelines into prompts.

The outcomes of the experiment were baffling and raised more questions. I am going to share the findings of this ground-breaking study with you.

In the ever-changing field of translation, achieving both efficiency and accuracy is vital. This presentation promotes an innovative collaboration between memoQ, Artificial Intelligence (AI), and Stream Deck. Together they form a powerful trio that enhances and streamlines a translator’s workflow.

At the heart of this partnership is memoQ, a robust, computer-assisted translation environment with a range of features. It ensures consistency through translation memories, manages terminology for accurate language use, supports collaboration on team projects, and allows for customisable workflows. In essence, memoQ is the key to increasing productivity, maintaining translation quality, and facilitating efficient collaboration.

The infusion of AI introduces new and enhanced capabilities such as alignment, term extraction, simplifying complex text for optimal use of machine translation (MT), clarifying poorly written source text, and providing thesaurus functionality. AI greatly enhances the translator’s toolkit, covering tasks such as writing regular expressions, checking grammar and spelling, translating poetry, and defining terms in context.

An essential part of this power trio, the Stream Deck orchestrates seamless interaction by providing customised shortcuts for quality assurance, translation environment functions, and speech-to-text transcription. Its utility extends to programming buttons with common steps, speeding up file management, and creating customised work profiles for different tools or clients.

The Stream Deck becomes a central hub for those navigating different translation environments, providing a consolidated platform for automating work and actions.In summary, the integration of memoQ, AI, and Stream Deck represents a paradigm shift in the translation landscape.

Beyond their capabilities, this synergistic harmony enhances the translator’s capabilities by providing a comprehensive and dynamic workflow solution. Through this collaborative approach, we envision a future where translation efforts are characterised by unparalleled efficiency, accuracy, and adaptability.

Working on large multilingual projects, with requirements for life sciences clients creates a unique set of process challenges.

To deal with these challenges Conversis has implemented many solutions using the memoQ API and project templates, as well as solutions using small tools developments (using regular expressions) and custom filters.

In this presentation Ian will show some of the major challenges faced and the solutions they implemented to solve them. With the hope that they may inspire you in your own endeavors.

In specialized fields like legal translation and EU institutional translation, the verification of the AI output based on reliable databases is an essential but often time-consuming task. The translated text should comply with requirements like consistency, accuracy, correct terminology and referencing, where the role of human expertise is unquestionable.

The online tool Juremy.com was designed to support precisely this workflow by providing instant contextual search results on large databases of EU documents. The tool brings fast bilingual segment hits in EUR-Lex and IATE in all 24 EU languages.

As a result of a recent development, an integration of Juremy and memoQ is already available for users. This new feature can significantly speed up terminology research, and thus increase productivity and translation quality.

In this presentation we will demonstrate practical examples on how the Juremy plugin speeds up terminology research by enabling memoQ users to carry out multilingual text searches in official EU resources.

Pseudo-translation means testing the translation process so that you can have an idea what will happen to your files.

Originally designed for software localization to make sure the translated text can be displayed correctly in the target language and fits the available space, pseudo-translation also helps to find hard coded text, i.e. text that the translation tool would not be able to extract for translation.

For other file formats, like PowerPoint, Word, InDesign, it can be used to find areas where non-importable text appears for example on images or in diagrams.

In addition, it can help to check your filter settings, especially XML and text-based filters, and make sure all translatable text is imported for translation and the text length settings (for text expansion) come in handy when you want to whether layout changes might be necessary after translation.

In this talk we hear from Benjamin Warren, Language & Translation Office Manager at Amadeus, who will share some insights from the exploration that he and his team undertook to examine the best course of action from various generative AI options at Amadeus.

You’ll hear more about what worked, and what didn’t, as they tried to understand how they could find real value from this much talked about technology.

We’ll look into the tests that were done, the results, and ultimately the choice that was made to deploy generative AI into their memoQ workflows.

Discover how they established the value that they wanted to find and how they measured its effectiveness as part of their program. We’ll look at the steps taken and how they made the journey from testing to production.

Those that attend this talk will come away with real-world examples of how generative AI can bring value to enterprise localisation programs and share some valuable ideas on where to start your own exploration.

In the Language Industry, the term ‘disruption’ is commonplace. We’ve heard it for decades and have become somewhat resistant to it. Despite this, we often consider the industry resilient, having weathered the global financial crisis and COVID-19. However, the advent of AI is raising new concerns. Will AI fundamentally alter our work, and will the profession of translation vanish? This presentation aims to address these questions and explore how AI will transform our lives forever.

Join us to explore different use-cases for the seamless integration of memoQ and Plunet, creating fully automated project roundtrips. Learn how to effortlessly set up projects, assign resources, and automate every workflow step. Say goodbye to manual interventions with automated quote creation, vendor requests, and assignments.

Discover Plunet’s flexible configuration options to tailor the perfect solution for your workflow. Witness the synergy between Plunet and memoQ, enhancing efficiency, reducing manual tasks, and ensuring high-quality outcomes.

memoQ is one of the very few translation environments that offers easy access to corpora – LiveDocs – that provide alignment functionality, but can also be used as a direct source of both mono- and bilingual reference for translation and concordancing. Alignment is the process of creating bilingual content from monolingual parallel documents.

Depending on the quality of the documents in question, the automatic alignment procedure can be fast and accurate or it can require a lot of time and effort. However, depending on the intended use of the aligned documents, quite often the exact, human-edited alignment is not necessary for useful results from rough, automatic-only alignment process.

LiveDocs can contain bilingual (or multilingual) resources in supported formats: it is possible to export translated, bilingual documents from memoQ projects directly to attached corpora and use them for matching and concordance in the same or other projects. This approach “freezes” the order of the segments (which is not guaranteed in a working TM) and make it possible to inspect matches in a visual context, if preview for a document was available.

Monolingual resources can also provide matches for source/target language concordancing.LiveDocs can also host any type of non-interpreted “binary” files, that the project users and download run locally. These can include, for example, PDF files, either reference source files or things like project instructions or client style guides.

Other examples include video as reference for subtitles localization, especially when combined with memoQ video preview tool.

The presentation will provide a quick overview of the LiveDocs corpora usage scenarios and how to get the most out of them.

This master class will feature functionalities that are available in memoQweb, the browser-based interface of memoQ TMS. Many of you have been using the desktop edition of the software for ages, and had no chance to see the improvements and the productivity boosters that will make the life of project managers way much easier.

This session will

• offer a discovery ride on memoQ’s web interface
• dig deep down in PM functionalities,
• cover Business Analytics, AIQE (Artificial Intelligence-based Quality Estimate, memoQ’s MT evaluation solution), In-Country review (ICR) tool and many other topics

Target audience: project managers | Level of memoQ knowledge required: basic/intermediate / advanced

Join memoQ’s Anna Mohácsi-Gorove, Inès Berbudeau and Luz Elena Videgaray Aguilar for this session!

Imagine a scenario where you are in the crossfire of clients and vendors: clients want to cut costs, vendors want to increase prices, and you are here to add value for both, while optimizing your workflow and making a profit as well. Sounds familiar, huh? Templates were developed to cut down the time and cost spent on translation project preparation. This session will feature the basics of translation project templates and explain how this resource can be used to automate and further optimize your translation workflow.

Target audience: project managers | Level of memoQ knowledge required: intermediate / advanced

The masterclass on TMS/server optimization at the memoQfest 2023 provided a good opportunity for advanced server users to meet and discuss various performance issues with memoQ’s developers.

The goal of this session would be to follow up on the issues discussed last year and provide a forum for server managers to share insights and discuss persistent challenges with the developers.

Highlights

• Presenting some practical tips and tricks that others might benefit from
• Discussing issues and requests with memoQ’s production team
• TM+
• Background tasks and queue revisited
• Questions about upcoming changes
• Discussion on missing features or features that need improvement

Target audience: server administrators, language engineers, power users

Introducing Breeze: A Breath of Fresh Air for Streamlining RFPs and Other Business Documents

Join us for a dynamic 25-minute presentation as we unveil Breeze, your ultimate solution for simplifying complex business documentation processes. Discover how Breeze revolutionizes the way you handle RFPs, but also Security, DDQ, DEI, and ESG questionnaires, saving you time, resources, and headaches. Learn how Breeze’s intuitive platform empowers you to breeze through paperwork with ease, unlocking newfound efficiency and boosting your organization’s success. Don’t miss out on the chance to breathe easier with Breeze!

MED-EL is a global hearing implant company with over 2,700 employees in 136 countries worldwide. We translate materials into around 40 languages daily. As a medical device company, we are highly regulated and need to follow strict standards regarding the quality of translations–and terminology.

Due to a lack of terminology management and no centralized translation department, many inconsistencies arose, which led to confusion for implant users and professionals. Our attempts to implement terminology management in the past were unsuccessful.

This time, we established a whole team to take on this project. We needed support from an expert, our long-term LSP, for training on how to implement terminology management in our company. Together, we came up with a structure for our new TB, selected the metadata for QTerm, and established a strategy.

Let us guide you through our first steps of the terminology process from merging several TBs to setting up a memoQ project in which we imported the QTerm termbase so we could benefit from existing TMs and TBs. This is where things got more challenging.

We encountered obstacles with term IDs not matching after reimporting, images disappearing when reimporting translated XLSX-files, and forbidden and alternative terms not showing up in the memoQ editor.

During our presentation, we will go into detail about how we were able to creatively overcome these challenges, use QTerm and memoQ to their full potential, and we will tell you about our key learnings.

With our presentation, we hope to show companies, LSPs, and freelancers the value of terminology management and how much can be achieved through the right strategies, good planning, and teamwork.

memoQ’s Customer Portal offers several benefits. Most valuable for us – automatic project creation. Leveraging this feature, we established a workflow that creates projects and assigns them to translators automatically.

Graphisoft offers its Subsidiaries and Partners the opportunity to participate in a (semi-)automatic process for translating marketing materials. In this process, a marketing professional from the Subsidiary orders the content relevant to them via the Customer Portal and the files are automatically assigned once the memoQ project is created. A dedicated team of in-house translators use memoQweb to translate the materials, which are automatically sent to the Customer Portal upon completion.

Involving in-house translators adds an extra layer of efficiency, as there is no need for further reviewing of the materials.Using this automated process significantly reduces the dependency on project managers.

Besides enhancing efficiency, this workflow also marks a shift towards a more automated future.

As a hype magnet the word “AI” often generates PR talk that’s long on vague promises but short on specifics. We at memoQ prefer no-nonsense solutions whose value can be demonstrated without buzzwords.

In this short talk memoQ AGT’s initiator will show how Adaptive Generative Translation works, how it fits into memoQ’s existing workflows, what we’ve learned in the first 6 months since AGT’s release, and what we’re working on next.

Caution! This talk includes hard numbers about quality.

Don’t get confused: the video below has been merged with the following session, that is why it’s called ‘Backstage Tour’.

Let the production team guide you through the most recent developments, and get a glimpse of what’s cooking at memoQ right now.

Don’t get confused: the video below has been merged with the presentation on memoQ AGT; that is the reason why it starts with Gabor Ugray’s presentation.