Skip to main content

Localization and LQA

Localization and LQA

This is for SaaS teams that already run machine translation, or want to, and need a person to make the output right before customers read it. We post-edit machine translation and run linguistic QA for documentation, UI strings and help centers in any language pair. Romanian and Russian are handled in-house at native level; other languages go to vetted partner linguists, and we manage quality from the first glossary to the final file.

Price

$0.05 per word; LQA $45/hour

Response

Reply within 1 business day

After delivery

30-day fix window

Who it's for#who-its-for

  • Heads of support launching a help center in a new market, who need translated articles customers trust enough not to open a ticket anyway.
  • Docs leads running Docusaurus or MkDocs i18n, who need translations that fit the repository workflow instead of spreadsheets emailed back and forth.
  • Product and engineering leads whose UI strings are machine-translated and need linguistic QA in context before a release goes out.
  • Localization managers who want an independent LQA pass on a vendor's work, with findings scored the same way every time.

When partner linguists are involved, you still have a single point of contact. We brief them, answer their questions, run the checks and deliver the files, so you never coordinate a chain of vendors. The service is built for product content, where accuracy and consistency matter most; it is not a fit for legal translation that needs a certified translator.

What you get#what-you-get

Depending on what you need, one or both of the following, each built around your terminology, because inconsistent terms are the fastest way to make translated docs feel untrustworthy.

MT post-editing:

  • Machine translation output edited to publishable quality: correct terminology, natural phrasing, a consistent tone, and formatting left intact.
  • Work done in your files and workflow: Markdown and MDX in Git, JSON or XLIFF string files, or your translation management system.
  • A glossary and style notes for each language, built from your product's terminology and kept current as the product changes.

Linguistic QA:

  • A review of existing translations, in context where possible, scored against an error typology such as MQM, with each finding categorized by type and severity.
  • A findings report with examples and corrections, so your vendor or team can fix patterns rather than chase single strings.
  • In-context checks of UI strings for truncation, broken placeholders, plural forms and locale formats for dates, numbers and currency.

Both come with a short delivery note per language: what changed, which terms were added to the glossary, and any source text that was ambiguous or wrong in English. Source problems are common, and fixing them once helps every language.

How it runs#how-it-runs

  1. Sample. You send a representative sample and the language pairs. We assess the machine translation quality and the content type, and confirm the rates and the linguists.
  2. Glossary. We build or review a glossary and style notes with your team before volume work starts, so terminology questions are settled once.
  3. Post-editing or LQA. Romanian and Russian are done in-house. Other languages go to partner linguists we have vetted, and every delivery passes the same checks for terminology, formatting and completeness before it reaches you.
  4. Delivery. Files come back in the format they arrived in, with a change log and any open questions for your team.
  5. 30-day fix window. For 30 days after delivery, any error you or your in-market reviewers find is corrected at no charge.

We work on the language side, not on engineering your i18n pipeline. If your docs site needs internationalization set up before translation can start, that is part of Docusaurus development, and the two can run together. When the product changes, only the changed pages or strings go back through post-editing; in a Git workflow, the diff shows exactly what moved.

Pricing#pricing

MT post-editing
$0.05 / word
about €0.05
Linguistic QA
$45 / hour
about €41.40

Post-editing is priced per source word and LQA by the hour, confirmed after we review your sample. The written quote lists the languages, the word count and the LQA hours, so the total is known before work starts.

What moves the total: the quality of the machine translation you start from, how technical the content is, whether a glossary already exists, how much in-context review the UI needs, and how many languages are in scope. Very weak machine translation is sometimes not worth post-editing at all; if the sample shows that, we tell you before quoting rather than after.

For teams that ship often, translated release notes can follow the English ones through the same review gate as part of release ops.

Prices are in US dollars; the euro figures are approximate conversions for reference.

Frequently asked questions#faq

Which languages do you cover?

Any language pair. Romanian and Russian we translate and review in-house at native level. Other languages are handled by vetted partner linguists who are native speakers of the target language, while we manage the project, the glossary and the QA.

What is MT post-editing?

Machine translation produces a first draft, and a linguist edits it until it reads as if it had been written in that language, with your terminology and tone. For most documentation it is faster and cheaper than translating from scratch, and the quality depends on the editing.

What is LQA?

Linguistic quality assurance: a structured review of translated content against a defined error typology, done by a linguist who did not produce the translation. It tells you whether a vendor's or a machine's output is good enough to publish, and which patterns to fix.

Can you work inside our Git repository?

Yes. We can work directly in Docusaurus or MkDocs i18n folders through pull requests, which keeps translations versioned alongside the source.

Do you localize UI strings as well as docs?

Yes, in JSON, XLIFF or other string formats, with in-context checks for truncation, placeholders and plural forms wherever we can see the running UI.

Send us what you have

Send a link to your docs, an export, or a few lines on what needs to change. We reply within 1 business day with questions or a quote.