Combining artificial and human intelligence meaningfully: Our services with AI

AI technologies are changing extremely quickly. As are most services with AI in the translation industry. These services are often also quite unique. However, there are some standards that we follow in our everyday work for our clients. You’ll find these established workflows below:

Where does Gemino currently use AI in its regular translation workflows?

We use the following workflows with AI regularly:

AI-assisted translation

• Workflow with the same steps as in the “traditional” process: Translation > Revision > Quality assurance
• In addition to hits from the translation memory (TM), translators are also supported with suggestions from an MT engine (NMT, LLM-based) during the translation step.
• The advantage: The use of existing translation memories is straightforward.
• The workflow is fully compliant with the requirements of ISO 17100.
• Savings: up to 15%* of the translation costs of “no matches” (i.e., new words to be translated)

 Verified AI translation

• Workflow with the steps: Machine/AI translation > Verification > Quality assurance
• The translation step is carried out by an MT engine (NMT, LLM-based).
• A translator then checks and verifies the machine translation (comparable to a revision according to ISO 17100). A final quality assurance step follows.
• This workflow does not involve a second translator and is therefore not ISO 17100 compliant.
• Savings: up to 35%* of the total translation costs

* depending on the text type and complexity of the specialist terminology

We always work with our clients to decide which AI translation technology to use. All framework conditions are taken into account as well as the suitability of the respective technology for the specific application. And, of course, as a matter of principle, we only use AI technologies in consultation with our clients.

Regular AI workflow for video localization

 We also use AI technologies for video localization. On the one hand, for the transcription of spoken language as the basis for a translation; on the other hand, for the creation of subtitles and speaker scripts. Speech synthesis is also increasingly being used as an alternative to voice recordings with voice actors.

Our conclusion

Artificial intelligence offers many possibilities for translations. At Gemino, we are following this development and helping to shape it. We use these technologies (where possible and sensible) in our regular workflows, but also flexibly and agilely in new services for our clients.

However, as tempting and simple as the use of AI may seem, careful evaluation and planning should always be carried out beforehand. This is the only way to actually reap the expected benefits.

Drawing on our experience 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. Because: Artificial intelligence still works best in combination with human intelligence.