Engine Suggestions used to come only from low-scoring AI reviews. Now they also learn from what human reviewers fix by hand. When a reviewer edits a translation – whether in-house or through an external provider – the engine compares the change against what it first produced, and when the same fix recurs across jobs or clearly sets a term, it proposes the glossary, instruction, or brand-voice edit behind it.
A fix a reviewer keeps making by hand used to be repeated work on every job. Now the engine surfaces the rule behind it: apply once, and the next translation already carries it.
Suggestions can now be approved or dismissed in bulk: select a set and act on all of them at once, instead of one card at a time. Engines with auto-approve on apply review-based suggestions on their own; the ones drawn from human edits always wait for manual approval.
