Most translation companies use off-the-shelf AI tools. Hansem Global built its own, extending what the industry-standard tools can do rather than replacing them.
Today, professional translation runs on CAT (computer-assisted translation) tools. Hansem Global, a localization company that produces and adapts industrial and technical documentation in more than 50 languages, works inside these platforms, and so do most of our clients and partners. Many of them now ship with built-in AI and machine translation.
But for high-stakes content like industrial equipment and medical device manuals, where accuracy and domain expertise are non-negotiable, the built-in AI couldn’t reach the quality bar we needed. And switching tools wasn’t a real option: it would mean migrating years of translation memory (TM), active projects, and the workflows our teams know by heart.
So we made a different call. Instead of changing tools, we built our own AI translation plugins that run inside the environments our teams already use. Rather than simply adopting AI, we engineered it ourselves.
Familiar Tools, Underpowered AI
memoQ and Trados are two of the most widely used CAT environments in the world, and both offer built-in AI integration. But for specialized technical content, that default AI revealed the same limitations:
- Limited model choice: You’re restricted to a narrow set of AI engines, making it hard to plug in the best, most current LLM at any given moment.
- Limited control: Tuning the AI to each client’s terminology and style guide, and to our own production workflow, simply isn’t possible with the default features.
A tool being familiar and the AI bundled with it being good enough are two very different things.
So We Built Our Own
Hansem Global developed its own AI translation plugins for both memoQ and Trados. Using each tool’s SDK, we connect today’s leading LLMs directly into the working environment. Translators get better AI output on the same screen they already use, whether they’re in memoQ or Trados.
memoQ and Trados are built on different architectures, yet Hansem Global applies the same approach to both. Whatever standard tool you use, we can layer better AI directly on top of it.
What Changes
- Direct integration of the latest LLMs: You’re not locked into default options. We connect the strongest model available at any given time, so your work always benefits from the latest performance.
- Tool-agnostic by design: Whether your team works in memoQ or Trados, you get the same improved AI inside that environment. No lock-in to a single tool.
- Glossaries and style guides that actually work: In technical translation, consistent application of glossaries and style guides makes or breaks AI quality. We feed client-provided glossaries and style guides into the AI translation step, and build them ourselves when a client doesn’t have them. The result: faster turnaround with consistent terminology and voice.
- Your workflow stays intact: Existing projects, TM, and review processes remain unchanged. You keep the stability of industry-standard tools while raising AI performance a level.
What’s Next: Translation You Control, Down to the Data
Right now, we connect the best available external LLMs to secure top-tier translation quality. The next step goes further.
Building on more than 20 years of domain- and language-specific data, Hansem Global is developing its own domain-adapted translation model to connect into these plugins. (For a deeper look at this topic, see “How to Go Beyond the Limits of General-Purpose AI Models in Translation”. Because it’s our own model rather than an external one, it makes the following possible:
- Data separation: Your data is never mixed into another client’s model in any identifiable form.
- Client-dedicated models: When needed, we build a dedicated model from your translation data, supplemented with synthetic data.
- Data control: Instead of depending on an external LLM provider’s policies, Hansem Global controls how the data is handled.
Because a custom, closed model is likely, especially early on, to translate at slightly lower quality than the latest general-purpose LLMs, there’s a genuine trade-off between cutting-edge LLM quality and full control over your data. Our goal isn’t to force one or the other, but to design that balance around each client’s content and security requirements. This is a longer-term initiative currently in the data-preparation stage, and we’ll share more as it matures.
Same Tools. Different Results.
Great translation doesn’t come from switching tools. It comes from combining better AI with human expertise, on top of the environment you already know. Hansem Global built exactly that, directly into the industry’s standard tools, and into both of them.
The most reliable way to see the difference is to see it. Run a sample of your own documents through Hansem Global’s AI translation plugin, inside the memoQ or Trados environment you already use, and compare the results. That’s not a vague promise but a difference you can see.
To request a pilot or a sample translation, contact Hansem Global.
Related Reading: How to Adopt AI Translation the Right Way
If this article was about how Hansem Global works with AI, the series below is about how companies should adopt AI translation.
- Part 1. Where Should You Start with AI Translation?
- Part 2. Why Do Results Differ So Much If Everyone Uses AI Translation?
- Part 3. How to Go Beyond the Limits of General-Purpose AI Models in Translation
- Part 4. Is AI Translation Effective for Every Project?
- Part 5. Where Do AI Translation Quality Failures Happen, and How Can You Prevent Them?
- Part 6. What You Must Check Before Adopting AI Translation
Frequently Asked Questions
Can Hansem Global add its own AI translation plugin to both memoQ and Trados?
Yes. Hansem Global built custom AI translation plugins for both memoQ and Trados using each tool’s SDK. Whichever environment your team works in, you get the same latest-LLM-powered AI translation inside it.
Is it better than the AI translation built into memoQ and Trados?
The built-in AI in memoQ and Trados is fine for general use, but for specialized technical manuals it’s limited in both model choice and control. Hansem Global’s plugin connects the latest LLMs directly and applies your glossary and style guide from the AI translation step onward, producing more consistent results.
Can you connect the latest LLMs (such as GPT or Claude) directly to translation work?
Yes. You’re not locked into default options; we plug in the strongest current LLM at any given time, so your work always benefits from the latest performance even as models evolve.
Can we keep our existing translation memory (TM) and active projects?
Yes. The plugin runs inside your existing memoQ or Trados environment, so your translation memory, active projects, and review workflows stay exactly as they are. There is no migration to a new tool and no need to move your assets.
What if we don’t have a glossary or style guide?
In technical translation, consistent glossary and style-guide application has a major impact on quality. If you provide them, we connect them directly; if you don’t, Hansem Global builds them and feeds them into the AI translation step.
How is data security handled with AI translation?
Today, because the plugin connects external LLMs, data handling follows the policy of the connected model. For clients who need a higher level of data control, Hansem Global is developing its own domain-adapted translation model. With that model, your data is never mixed into another client’s model in identifiable form, and Hansem Global controls how the data is handled.