Import from Legacy Vendors
When migrating from a legacy localization vendor or TMS, you likely have glossaries, term bases, or translation memory exports sitting in TMX, CSV, or TBX files. Your AI assistant can parse these and seed your localization engine's configuration directly.
The workflow#
"Here's our glossary export (CSV). Import it into our engine's glossary for all locales."
What happens:
- The assistant reads the CSV structure — identifies source term, target localization, locale, and term type columns
- Groups the rows by source term. A glossary term is one source text with its localizations for every target locale; a non-translatable term has none
- If your engine applies more than one glossary, asks which one to import into
- Compares the export with the terms the glossary already holds and shows the import plan: "Found 140 rows describing 25 terms across 6 locales. 2 are do-not-translate, 23 are enforced localizations. 4 of them are already in the glossary with different localizations — update them?"
- On approval, creates the new terms and updates the existing ones you confirmed
- Reports: "21 terms created, 4 updated."
A term the glossary already holds with localizations is never overwritten by an import. The assistant updates it only when you confirm, and keeps the localizations for locales your export doesn't cover.
Supported formats#
| Format | What it contains | How to provide it |
|---|---|---|
| CSV / TSV | Term bases, glossaries, simple bilingual lists | Paste the content or describe the file structure |
| TMX | Translation memory — source/target segment pairs with metadata | Paste a representative sample or describe the structure |
| TBX | Terminology databases — structured term entries with definitions | Paste the content or describe the schema |
| Excel exports | Vendor-specific glossary or style guide exports | Describe the columns and paste representative rows |
Step-by-step: TMX import#
TMX files from legacy vendors contain segment pairs that can seed both glossary entries and rules.
"Here's a TMX export from our previous vendor. It has 500 translation units for en → de. Extract any recurring terminology as glossary entries."
What happens:
- The assistant parses the TMX structure — identifies source segments, target segments, locale pairs
- Groups recurring terms — words or phrases that appear 3+ times with consistent localizations
- Proposes glossary entries for terms with stable localizations: "'privacy policy' → 'Datenschutzerklärung' (appears 12 times, always localized this way)"
- Identifies patterns that should become rules: "Compound nouns are always hyphenated in this corpus — add as rule for
de?" - Checks the proposed terms against the glossary. A term already there with a different German localization is shown side by side: "'checkout' is 'Kasse' in the glossary, 'Bezahlvorgang' in the TMX — keep or replace?"
- Shows the full plan for review
- On approval, creates the new terms and rules, and changes only the existing terms you chose to replace
Step-by-step: CSV glossary import#
Most legacy localization platforms export glossaries as CSV with columns for source, target, locale, and notes.
"Import this CSV into our engine. Columns: source_term, target_term, locale, type (localize/do-not-translate), notes."
What happens:
- The assistant reads the column mapping
- Creates glossary entries:
localizerows become custom localizations,do-not-translaterows become non-translatable entries - Entries with notes that describe rules (not just definitions) are flagged as potential rules: "The note for 'date format' says 'Always use DD.MM.YYYY in German' — add as rule for
de?" - Shows the plan, applies on approval
What to import vs. what to leave behind#
| Import as glossary | Import as rule | Skip |
|---|---|---|
| Brand names (non-translatable) | Formatting rules (date, number, currency) | Fuzzy TM matches below 95% |
| Product terminology (enforced localizations) | Punctuation conventions | Context-dependent segment pairs |
| Legal terms (enforced localizations) | Register/formality rules | One-off localizations that aren't terminology |
| UI labels with mandated localizations | Capitalization rules | Segments longer than 2-3 sentences |
After import#
- Verify — localize content that uses the imported terms through the localization API or the CLI
- Review — spot-check a batch against the new glossary to confirm enforcement
- Tune — adjust entries that don't produce the right output in context
Tips for large imports#
- Start with high-frequency terms. A 5,000-entry TM export isn't a glossary — it's a corpus. Ask the assistant to extract only terms that appear 3+ times.
- Review by locale, import all locales together. Reviewing 50 German terms is easier than 500 terms across 10 locales, so go through the plan one locale at a time. Then import every locale in one pass: a glossary term holds all its localizations, so a locale added later changes terms that already exist instead of adding new ones, and each of those changes needs your confirmation.
- Expect large imports to run in parts. The assistant writes up to 100 terms per call. Each call is all or nothing, so if one is rejected, the assistant reports which terms caused it. Calls that already succeeded stay written.
- Use the notes column. If your export has translator notes, the assistant can convert patterns into rules.
- Don't import sentence-level TM as glossary. Glossary entries are terms and short phrases. Full sentences belong in reference material, not the glossary.