GTranslate Pro keeps its automatic and human-corrected translations in a dedicated gt_translations table, with source string, target language, translated value, and an edited flag. Its default admin shows a single string editor with a search box, which works for a quick fix but not for a weekly audit of 60,000 cached strings across eight languages.

SleekView reads gt_translations alongside wp_options entries like GTranslate config and treats every language as a first-class column. Source string, original URL, translation status, and last-edited timestamp become sortable, filterable fields. Saved views like edited-this-quarter or untouched-machine-translation-in-French load with one click.

Edits route through GTranslate's own translation API where available; for direct adjustments to the cache, SleekView writes back to gt_translations with a soft optimistic lock so two editors don't quietly overwrite each other.

GTranslate Pro is designed for fast multilingual rollout: configure a few languages, point it at a translation engine, and your store renders in eight locales the next day. That speed is the selling point, but it leaves a long tail of machine output sitting in gt_translations that nobody has reviewed. The default string editor handles the urgent fix well enough, but it cannot answer questions like which 4,000 strings are still raw machine output in French, or which buttons changed in English last month and therefore need a re-review across every language.

Treating the cache as a queryable table changes that posture from reactive to proactive. A localization manager can see at a glance how much human-reviewed coverage exists per language, prioritise the strings that drive conversion, and hand reviewers a scoped CSV instead of the full translation cache. Inline edits still route through the same table that GTranslate Pro reads at the front end, so corrections appear on the next page render with no extra deployment step.

The result is that machine translation becomes a starting point rather than an end state, and the audit fits inside an existing weekly review rather than a separate localization tooling stack.