The Localization Project Manager checks the release calendar for a product version or feature branch and:

1. Optionally, verifies the zTranslate.ai Product Settings & Assets is accurate from last use:
   a. The supported languages in the Product Details
   b. The TM assets for all supported languages are updated in the central Product Translation Memory repository
   c. The context assets from previous projects have been saved to the central Product Context Screenshot repository
   d. The users and vendors needed for the localization cycle have been included with their roles accurately specified in the Product Team
   e. The automation settings to external tools such as GitHub, JIRA, Worldserver, and Machine Translation site are working in the Product External Integrations

2. Creates a zTranslate.ai Project in the core product language, say English, and assigns it to the Localization Engineer
   a. zTranslate.ai notifies the Localization Engineer by email including a link to the project

Once a Localization Engineer receives notification that a new project is ready to localize product resources, they go through a sequence of steps:

1. The Localization Engineer uses the Import Resources project action to load the localizable product string resources from the source repository, such as Github:
1. zTranslate.ai creates tasks in the English project for each of the resource files that are imported.
2. zTranslate.ai creates localized sub-projects for each supported language, and copies the resource file tasks.
3. zTranslate.ai leverages the applicable Product TM on the resource tasks of each project.

2. Assigns each of the language projects to Localization QA
1. zTranslate.ai notifies the Localization QA by email including a link to the project

When the QA Engineer executes their localization test suite to test the product, they will incorporate some additional actions in their workflow:

1. Use the zTranslate.ai Context Kit Creator (CKC) Chrome extension to take product screenshots in English, and repeat while running the app in each supported language
1. CKC builds up the product context data
2. highlights the strings on all screenshots
3. links strings to screenshots
2. Use the ‘Send to zTranslate.ai’ button when done with all screenshots
1. zTranslate.ai updates the English project with the provided context info
2. zTranslate.ai updates the supported languages with context info
3. Assigns each of the language projects to a Translator or Reviewer Linguist
1. zTranslate.ai notifies the Linguist by email including a link to the project

Once a project is prepared the Localization engineer needs to communicate with the linguists - first the translators to notify them that a translation project is ready, and to prepare a translation kit to send to them to use in their preferred translation tool

1. The Translator uses the Export Translation Kit project actions to download translation kits to work in preferred tools like Trados, Passolo etc. and Import Translation Kit to update the translations.
2. The translator assigns the project to the Linguistic Reviewer who receives an email notification from zTranslate.ai.
3. Reviewer verifies and validates all strings, and corrects translations that are contextually incorrect
   a. zTranslate.ai enables in-house translators and smaller projects to be translated and reviewed directly in the Linguistic Workbench.
   b. zTranslate.ai enables the Reviewer to mark a string as validated.
   c. zTranslate.ai enables the Reviewer to look up all product context screenshots for a string.
   d. zTranslate.ai enables the Reviewer to update translations and mark it as validated.
4. Reviewer assigns the project to the Localization Engineer and zTranslate.ai sends an email notification

Once a project is prepared, the Localization engineer needs to communicate with the Vendor manager regarding the translations that receive an email notification from zTranslate.ai.

1. Once the vendor manager has received the project, they will attach the translation quote to it and return it to the project manager.
2. After reviewing the quote, the project manager will return it to the vendor manager only if it is approved.
3. After the quote is approved, the vendor manager sends the project to the translator for translation.
   a. The Translator uses the Export Translation Kit project actions to download translation kits to work in preferred tools like Trados, Passolo etc. and Import Translation Kit to update the translations.
   b. The translator assigns the project to the Vendor Manager who receives an email notification from zTranslate.ai.
4. Vendor Manager reviews the project and then assigns the project to the Localization Engineer and zTranslate.ai sends an email notification.

The Localization Engineer uses the Create Target Project project action to create all target projects in one shot, or the higher level processing shown in the localizable resources section.

1. zTranslate.ai creates all target locale projects in one shot

2. zTranslate.ai copies all resource tasks from the source language project to each target language project.
3. zTranslate.ai leverages from the respective language central Product TMs if the user has selected the Leverage Product TM option.
4. When projects are assigned, zTranslate.ai notifies all translators by email. When translations and reviews are completed, the assigned Localization Engineer is notified by email.
5. zTranslate.ai provides the Export Resources project action to process all target projects and regenerate the translated resource files.

6. The remaining clean up tasks can be performed by the Localization Engineer, but are described in the Save Incremental Work section for the Localization Project manager, as they have the overall view of multiple related projects that are in progress over multiple code branches and sprints.

zTranslate.ai project action View Tasks shows each resource file as a task with counts: total strings, translated, untranslated, validated, unvalidated, and corrected

1. a project - as it flows through leverage, translation, and reviews for all locales
2. a product version - as each sprint or feature branch is processed multiple times
3. all product versions - the contribution of each product version by locale
4. zTranslate.ai project action View Tasksshows each resource file as a task with counts: total strings, translated, untranslated, validated, unvalidated, and corrected

5. zTranslate.ai project action View Reports allows all users to view rollup reports of a number of string statistics - by project state, by project locale, by product version, and click through links to view the filtered projects by state

The Localization Project Manager reviews project completion statistics to confirm a project has been completed.

1. zTranslate.ai enables the PM to do various closing activities via project actions:
   a. update all validated translations to the central Product TM repository using Update Product TM
   b. update the collected context screenshots for the project to the central Product Context repository using Update Product Context
   c. push final Product TM to Worldserver or other external translation application using Export Product TM
2. The PM can Close Project or Archive Project once all work is completed!

The UX designer or Product owner decides to use the English (Neutral) locale as their target language so they can maintain a change memory of their changes much like a translator uses a translation memory.

1. Creates a new project for a particular product release with their source locale and the Neutral target locale, and imports the English resource files.
2. Assign to a QA user who can create any additional feature screenshots since the last update to the central Product Context repository.
3. Start working making corrections in the Workbench searching for strings the have identified to change and viewing screenshots where context is needed. A second reviewer or the UX designer can mark the corrections as validated.
4. When done, the project is assigned to a L10N engineer who will use zTranslate.ai's Export Resources to commit into the source repository with the corrections. The Neutral locale TM is also updated with all validated changes using zTranslate.ai's 'Save Product TM'
5. The product owner can monitor their changes in the next product build, and return to Step 1 and repeat the process. zTranslate.ai provides a filter to exclude validated strings so the designer can focus on new corrections.