Still finding quality problems only after delivery?
Quality varies by translator, and checking after delivery doubles the work when files come back. tralio checks quality in real time while translators work, preventing problems before they happen.
Problems found after delivery
- ✕Each translator follows a different quality bar
- ✕Line-length and reading-speed (CPS) violations found after delivery
- ✕Inconsistent terms noticed only at review
- ✕Endless send-back, fix and re-check cycles
- ✕Quality standards passed on by word of mouth
Quality built into the system
- Line length, reading speed (CPS) and line count checked in real time
- Accurate Thai character counts that exclude tone marks
- Glossary AI flags term mismatches automatically
- Automatic checker assignment automates the review step too
- Standards live in the system, not in individuals' heads
How it works
Automatic QC
Characters per line, characters per second (CPS) and line limits are checked in real time as translators type. For Thai, a dedicated counter excludes tone marks and similar diacritics. QC standards adapt to each language automatically. Example: Japanese 39 characters/line, Thai 39 characters/line (excluding diacritics)
Glossary AI integrated with QC
When you register a term, AI detects its part of speech (noun / verb / adjective) and domain. Translations are matched against the glossary and mismatches are flagged automatically. Glossaries are managed per language pair.
Automatic checker assignment and line-by-line comments
Once the translator accepts, the checker offer starts automatically. Checkers leave specific feedback as comments on each line, shortening the fix cycle between translator and checker.
QC for multiple languages
Language-specific QC standards apply automatically — such as Japanese line limits and Thai counts that exclude diacritics. QC rules for new languages can be added flexibly.