Manage customized MT engine presets

The customized MT engine preset is a robust and convenient method for maintaining and managing the credentials of various MT developers.

You can add an API key for an MT developer and then use this key to create one or several customized MT engine presets. These presets will be available for addition to the Machine Translation section of the project settings together with the standard MT engine profiles.

To add an API key for an MT developer:

1. Log in to Smartcat as an account administrator.

2. Navigate to Settings > Credentials, click Add Credential and select the required MT developer. In the example below, the selected MT developer is Google Advanced Translation. 

3. Enter a meaningful Name, that you will later require to create a customized MT profile.

4. Optionally, enter additional information about the provider to the Comment field.

5. Paste the MT developer's API key to the Service account key field. 
   Note that For Google Advanced Translation the API key is provided in a JSON file. Open this file in any text editor, copy its entire content, and paste it to the Service account key field.

6. Click Save. 

1. Navigate to Settings > MT engine presets, click Add Preset and choose the same MT developer you selected at step 4 of the previous procedure. The Add MT Engine Preset dialog is displayed: 

2. Click the arrow to expand the API key drop-down list and the name you assigned to the MT developer at step 5 of the previous procedure.

3. Optionally, enter additional information about the preset to the Comment field.

4. If you are creating a preset for Google Neural Machine Translation or Google, click Save and skip the rest of the procedure.
   If you are creating a preset for Intento, clicking Save at this step will create a fully functional preset, but you can enrich it by providing a Routing name, that is the name of a strategy that Intento uses to customize the behavior of their MT output. 
   If you are creating a preset for Google Advanced Translation, clicking Save at this stage will create an operational MT engine preset that you can enrich with language specific parameters. To do it, click Add language pair.

5. Select the Source and Target languages for the preset.

6.  Optionally, use the AutoML Model ID parameter to provide the identifier of your trained AutoML model. To obtain this identifier, log in to your Google Cloud Console and follow these instructions.

7. Another optional parameter is Glossary ID, the identifier of your glossary to be used together with the trained AutoML model. For the information on creating and using glossaries in Google Advanced Translation, refer to this article.

8. You can also provide labels (key-value pairs) used in Google Advanced Translation for filtering billing charges in the Google Cloud console. To create a new label, click Add Label and specify the key-value pair. Keep clicking Add label to create as many labels as required.

9. Click Save when done.

To use a customized MT engine preset, add it to your project by clicking Add in the Machine Translation dialog, as described here.