The scorecard turns your quality findings into per-language scores in the vocabulary buyers and vendors already use.
Who this is for
Anyone who has to report on translation quality, and anyone being asked to.
What happens next
Per-language quality lists each language with its score, described on the page as "Numeric MQM scores in the DQF-MQM vocabulary enterprises RFP for: a deterministic transform over the QA findings, no extra AI cost."
Two things follow from that sentence and are worth knowing:
It is a transform, not a second opinion. The scorecard re-expresses findings QA already produced, grouped into the standard MQM categories: accuracy, fluency, terminology, style, locale convention. It does not re-examine your translations.
It costs nothing extra. Looking at the scorecard does not spend any of your AI allowance.
Use it to compare languages against each other, to see which category a language's problems fall into, and to answer a procurement questionnaire without hand-assembling numbers.
If it doesn't work
"No scored languages yet" / "Run QA on this project to populate the per-language MQM scorecard.": nothing has been checked yet.
Related
Quality checks
What does the quality score mean?