Compare
The platform handles the content. The document is a different problem.
Software localisation and content localisation are well served by this category. Documents are the case that keeps falling outside it.
What it is good at
String and content localisation is the core strength. Keys, files, continuous delivery — the software side of localisation is where this category is built to operate.
Repository and pipeline integration is mature. Content moves through automation rather than through email attachments.
The model is built around translatable units rather than around documents. That is the right model for software and the wrong one for a 60-page brochure.
Where the document itself is left to you
These are the gaps our pipeline exists to fill. They are about the file, not about the tool.
Documents are not strings
A catalogue or a slide deck has geometry, not just text. Localising the text and preserving the geometry are different problems.
Design files are outside the model
Text inside a designed frame has to fit that frame after translation. That is a typographic problem, and it sits outside a string-based model.
Scanned material has no keys
There is nothing to look up when the content only exists as pixels.
The visual result is unverified
A string can be translated correctly and still produce a document that looks broken. That failure is invisible until someone opens the file.
Capability comparison
Categories, not vendors. We do not publish anyone's pricing, including our own.
| Row label | Us | Raw MT / cloud translation APIs | TMS platforms |
|---|---|---|---|
| Multilingual | Configured to your markets | A fixed list | A fixed list |
| Format preservation | Layout written back in place | Text or a rough file | Not an engine |
| Per-page processing plan | Each page routed on its own | One call, no quality control | Manual workflow |
| Multimodal input | Five source classes, five paths | Mostly text only | No |
| What you receive | Finished files | A block of translated text | Process management |
| Who owns the quality | Us | You | You |
How this sits next to Phrase
Different content types, different paths, one agency delivering both.
The common arrangement: content and software localisation continue through the platform, and documents come through us. Both return finished output to the same client.
Where the two need to talk — a document that originates in the platform's content system, for example — that connection is scoped rather than assumed.
This is not a competition between categories. An agency serving a software client usually needs both, and they are not substitutes.
Questions
Is this an alternative to Phrase?
For documents, it addresses the part that content platforms do not cover. For software and content localisation, the platform is the right tool and we are not trying to replace it.
Can documents come from our content system?
That connection can be scoped. Tell us what your system can call and we will tell you what is workable.
Do you handle design files?
Not the native file. InDesign, IDML, Quark, Illustrator and FrameMaker are not formats our pipeline processes. A design exported to PDF or supplied as images we can take.
What about scanned documents?
They go down the optical recognition path. A PDF that mixes digital and scanned pages is handled page by page.
How is quality verified?
The output is reviewed by a separate model from the one that produced the translation, and the rendered result is compared against the source before delivery.
Send us a document and see what comes back.
The gap is easiest to judge with a finished file in front of you.
- No signup, no self-service portal — you send a file, we run it.
- You get the finished document back, not a screenshot of one.
- If it is not good enough, you have lost nothing.