AI is constantly changing. And currently extremely fast. Which is why most language services using AI technology currently still have a relatively short half-life and can or must be conceived very differently depending on the given framework conditions.
With this in mind, it would not be enough to simply list our current AI services here. Instead, we would like to present the topic of “AI and translation processes” comprehensively – with regularly updated information and specific tips on the optimal use of AI. This also includes application scenarios describing how we use AI in projects.
Updated 2025-06-25
Since the introduction of ChatGPT – probably the best-known Large Language Model (LLM) – artificial intelligence and the expectations associated with it have been omnipresent. At Gemino, we have since constantly been dealing with evaluating suitable areas of application and are already using AI specifically in project processes.
What we can say with certainty: as tempting as the use of AI may seem, it should always be preceded by careful evaluation and planning. This is the only way to reap the expected benefits and realistically assess the risks.
With “AI” or “GenAI” (Generative AI), we almost always refer to Large Language Models (LLMs) in the context of textual content. The best known include GPT (OpenAI), Gemini (Google), Claude (Anthropic) and LLaMA (Meta). In addition, neural machine translation (NMT) is the most important and currently (still) predominantly used technology. This is based on deep learning and – like LLMs – belongs to the “artificial intelligence” category. Other types of machine translation systems, such as statistical and rule-based systems, hardly play a role today.
NMT systems are based on neural networks. They are specifically trained for translating text with suitable corpora. LLMs, on the other hand, are developed for generating texts. These texts are then used to generate texts in another language. The distinct methodologies impart unique characteristics to their “translations”, each presenting its own set of advantages and disadvantages – much like any technology. This is why LLMs and NMT are suitable for different use cases.
When planning translation processes integrating AI systems, there are various aspects that should be taken into account:
Welche Anforderungen/Erwartungen beschreiben die Qualität, die erzielt werden soll? Bleiben die Vorgaben gegenüber den bisherigen Festlegungen unverändert? Welche Abstriche sind möglich und/oder akzeptabel? Auch das Thema Risiken ist wesentlicher Teil des Qualitätsbegriffs und sollte immer bedacht werden – besonders in Bereichen, die rechtlichen/regulatorischen Anforderungen entsprechen müssen (z. B. Medizinprodukteverordnung (MDR/IVDR) oder EU-Maschinenverordnung).
Do the processes and systems currently in use allow the integration of AI systems? What conditions need to be created? Do the achievable benefits justify changes to existing, established processes and the associated investments, which can be considerable?
What data protection and IT security requirements must be met? Is the use of cloud-based systems conceivable? These are almost always superior to systems used in your own data center in terms of technology and training quality. However, they must also comply with data protection requirements and your own IT guidelines.
The regulatory and legal requirements of the respective industry must be taken into account. For example, in the area of medical products: MDR and ISO 13485 require software to be validated if it is used in the production of medical devices or has a process-related influence on product quality. For a successful validation, however, it must be proven that the software delivers a defined output for a defined input. The software must therefore be able to generate the desired result reliably. To the best of our knowledge, this is not yet possible in the case of LLMs (keyword “hallucination”). However, this does not mean that AI systems cannot be meaningfully integrated into translation processes in this environment other ways.
One thing is clear: The appeal of AI lies in potential cost savings and faster production times.
But we recommend checking in advance whether the desired benefits can actually be achieved under real conditions (with all the special cases to be expected), if possible by means of a real proof of concept.
This means in practice:
What type of content is affected if a translation process switches to AI? Different types of content, i.e. text text types, work differently with AI. Also, the choice of the respective technology can play a role (LLM vs NMT system, see above). And there are significant differences in quality not only by system, but also by target language. And not all target languages are available in all systems.
So, if many small translation projects are to be processed, and different workflows with different systems are required for each language combination and content type, you need to take a closer look. This can result in only minimal time and cost savings compared to the traditional translation process – in the worst case, it can even lead to additional costs. Then, a traditional workflow variant may be more suitable than a largely automated workflow in which the quality is primarily characterized by the AI translation.
First and foremost, of course, AI is used for machine translation. In addition to the now established NMT systems, Large Language Models (LLMs) are an interesting and, depending on the application and general conditions, possibly the better alternative. For example, reference texts can be integrated in certain configurations with LLMs (through prompt augmentation, RAG, etc.). As a result, the AI-generated translation takes the desired reference into account and the output is adapted to linguistic specifications such as style or terminology.
One area of application that we are currently investigating is AI-supported quality assurance of translations. LLMs can support quality assurance through new types of testing that were previously not feasible in terms of time and budget. Even if fully automated checking and correction is not yet possible, we are currently working on setting up a system to support our experienced QA linguists. The use of AI for semi-automatic checking/correction or as a warning system for certain types of error looks very promising at the current stage of the project.
The so-called “agentic AI” is currently opening up additional application possibilities. Various AI systems are combined in workflows to form AI teams. This enables a much wider range of possible uses for AI – especially for multi-stage AI-based workflows that would be impossible to implement with a single system. One example: After the initial translation by an LLM, the text passes through a second LLM system. In this subsequent step, only the terminology is checked to the relevant specifications and definitions.
In addition to various new use cases for which we are currently still testing the use of AI and developing special solutions, we also use artificial intelligence productively as standard. Important here: we use the technologies exclusively in consideration of the aforementioned aspects and always in consultation with our clients.
We currently offer several translation workflows with AI. We work with our clients to decide which workflow is best suited and which machine translation technology to use. The evaluation takes into account all relevant external (technical) aspects as well as the suitability of the respective translation technology for the specific type of content.
For a quick overview, you will find below a table of all workflows with information on the services included, the potential savings and the respective compliance and quality aspects:
* Critical in areas where ISO 13485 is applicable: Topic “software validation” and topic “consistency with previous version for updates”. Conformity can still be achieved with the help of custom workflows. Individual consultation required.
** Custom workflows available for incorporating TM content.
Please note:
Not all workflows are suitable or applicable for all content types. During the project analysis, Gemino checks whether the content is suitable for various workflows. This ensures feasibility and the achievement of reliable results in accordance with the intended use, and that quality requirements are always guaranteed.
The two workflows “AI-assisted translation” and “Verified AI translation” in particular can be useful updates to the traditional translation process:
We have also been using AI technology for the optimized localization of videos for some time now. On the one hand, for the transcription of audio content as the basis for a translation; on the other hand, for the creation of subtitles or speaker scripts. Speech synthesis is also increasingly being used as an alternative to (extremely costly) human voice recordings.
AI offers many new opportunities in the field of translation and multilingualism – almost every day. At Gemino, we are following these developments and seek to influence them. Our aim is to have a realistic picture of the opportunities and risks at all times so that we can integrate new technologies into our processes – and those of our clients – in a meaningful way at all times.
What is certain: There are no fixed or “best-in-class” solutions. Every use of AI is individual and should be conceived to be agile so that its use can be adapted or expanded as required. Also so that new technologies can be easily integrated at any time.
When planning and implementing new multilingual processes with AI, the involvement of several specialist disciplines in the company is required – especially from a linguistic and translation logistics perspective.
Drawing on our experience with tools and processes in the field of multilingualism and our expertise in AI, LLMs, GenAI and the like, we are happy to accompany you on your way to new, better, and individual solutions.
In doing so, we follow the guiding principle: artificial intelligence still works best with the targeted addition of human intelligence!