Loqalit checks your signup, onboarding, pricing and in-app pages in every language you ship, on a schedule and before each release. It flags unfilled placeholders, buttons and menu items whose text no longer fits, English strings from new features and terms that drift from your glossary, each with a screenshot and a suggested fix. A GitHub Actions step can stop a release when a critical or major finding appears on staging.
Every release adds strings. The ones that slip through show up where users decide whether to stay.
The first screens a new user reads, where a broken sentence costs a trial.
Plan names, limits and prices that must be right in every market.
Menus, settings, dashboards, empty states and error messages.
Fresh strings that were never sent for translation, or came back late.
Articles and in-product help that should match the UI they describe.
In-app banners and notices that combine placeholders with grammar.
Every finding comes with a screenshot of where it appears, an MQM severity and a suggested fix that fits the space.
Add Loqalit as a GitHub Actions step pointed at your staging site, and critical or major findings fail the build.
Daily, weekly, monthly, quarterly or once, in the languages you choose, with nobody at the keyboard.
Scheduled scans read pages without logging in. Signed-in screens can be checked with the Chrome extension while you’re signed in.
Loqalit checks the product your users see. It doesn’t replace pseudo-localization, unit tests or your TMS’s file checks, and it doesn’t edit your code or your strings. Your TMS stays the system of record: accepted fixes can be posted to Crowdin or Phrase as comments with a screenshot.
Scheduled scans and the release gate read pages that open without signing in, such as signup, pricing and help pages. Signed-in screens such as settings and dashboards can be checked with the Chrome extension while you’re signed in.
Yes. Add Loqalit as a GitHub Actions step pointed at your staging site, and critical or major findings fail the build.
Accepted fixes can be posted to Crowdin or Phrase as comments with a screenshot. Findings can also be exported as PDF, Excel, CSV or JSON, or sent to your own tools through the API and webhooks.
Pseudo-localization tests your code before translation. Loqalit checks the real translations on the rendered product, so it finds wrong meanings, untranslated strings and real text that no longer fits.
Tell us which pages matter most and we’ll start there.