Below are the end‑to‑end procedures to deploy Globant Enterprise AI for Version 2025-11 and its Hot Fixes.
Check also version updates for the main platform components.
No manual steps required.
No manual steps required.
- Data migration (normalize file references in existing conversations)
- This migration updates existing conversation messages to use normalized file references.
- A Super Admin must run it manually from the Console > System > Metadata > AI Station > Conversation Message File FileId Fix > Apply FileId Fix > Click Confirm.
- To review results, go to System > Operation Logs.
- Notes:
- The migration only applies to existing file references. It does not apply to temporary files.
- The log can show files not found. This does not necessarily indicate a problem if those references no longer exist.
- The process runs in a submitted thread, so it is not a deployment prerequisite and you do not need to wait for it to finish.
- Verify Japanese characters (System module revision 27)
- After updating the System module to revision 27, verify the BACKOFFICE_AVAILABLE_LANGUAGES parameter to ensure Japanese characters display correctly.
- If the parameter already existed before the update, no action is required because the update does not overwrite it.
- If the parameter was introduced with revision 27, update the Japanese entry as follows:
Replace:
{"name":"Japanese","iso_code_6391":"ja","description":"???"},
With:
{"name":"Japanese","iso_code_6391":"ja","description":"日本語"},
This issue occurs because special characters are not mapped correctly in the DataProvider where the parameter is defined.
- Cost recalculation in Requests (from December 1, 2025)
- Recalculate request costs starting on December 1, 2025 for the following models:
- gpt-5-codex
- gpt-5-chat-latest
- gpt-5.1-codex-max
- gpt-5-pro
- gpt-5.2-pro
- You must complete two stages:
- Stage 1: Recalculate costs for each model
- Go to Console > System > Metadata > Request utils > Request cost recalculation.
- Set From to December 1, 2025.
- Set To to the current date.
- In Request AIModel Name, enter one model from the list above.
- Click EXECUTE COST RECALCULATION (SUBMIT).
- Check results in System > Operation Logs.
This process runs in a submitted thread.
Repeat steps 6 to 7 five times, once per model.
- Stage 2: Update dashboard summary data
- Go to Console > System > Metadata > Request utils > Reprocess Requests and load summary table.
- Click Load a specific period.
- Enter December 1, 2025 as the start date and the current date as the end date.
- Click REPROCESS REQUESTS (SUBMIT).
- Check results in System > Operation Logs.
This process also runs in a submitted thread.
No manual steps required. This change allows proper configuration of INSTALLATION_CONFIG_PROVIDER=geai for the Workspace when it is set.
- Flows SSO
- In GAM UI go to Applications > Fluentlab.
- Edit the configuration to add:
- Additional user scopes: openid
- Flows SSO in SaaS for clients with custom IdP and subdomain
- This is only for clients using the SaaS with their own IdP for login and custom subdomain.
- Assuming that the client is already setup to access Flow from GEAI Console.
- Go to Console > System > Parameters.
- Look for the client Organization.
- Edit the parameter UIConfiguration.
- In the JSON (Value S), under the customization attribute, add a new property called flows with the following content:
{
...
"customization": {
...
"flows": {
"idp": ""
}
}
}
The client IdP should match the one defined in the Keycloak from Flows.
- Reinstall system module “The Lab Public Components” to update Iris definition
- Go to Console > System > System Modules.
- Search for “The Lab Public Components”.
- Click REINSTALL/UPDATE.
- Do the same for Module: “The Lab Tools: Google Drive Tools”.
- Configuration for Lab, Station and Workspace
- The following configuration applies to the new header menu and help menu options.
- Define the following environment variable for Lab, Station and Workspace backends:
INSTALLATION_CONFIG_PROVIDER=geai.
- This defines that the next UIConfiguration parameter will be used to load the configuration defined in it.
- Go to Console > System > Parameters.
- Create a new parameter called UIConfiguration with the following content:
{
"configurationDomain": "<tenant_domain>",
"gamAuthTypes": [
"<authentication_name>"
],
"customization": {
"applicationsUrl": {
"station": "<station_url>",
"console": "<console_url>",
"lab": "<lab_url>",
"workspace": "<workspace_url>"
},
"support": {
"contactUs": "<contact_us_url>",
"helpUrl": "https://docs.globant.ai",
"tutorialVideo": "https://www.youtube.com/embed/0eXMmPRGXS4"
}
}
}
Where:
- <tenant_domain>: domain where the workspace is running, for example, customer-workspace.domain.com.
- gamAuthTypes is a list of Authentication name defined in the Security configuration.
- <contact_us_url>: for example mailto:support@domain.com
- Check that the parameter has the scope on System in Advance section.
- For the Station and Workspace you also need to set these environment variables in the frontend containers:
- VITE_GEAI_STATION_URL_VALUE=<station_url>
- VITE_GEAI_LAB_URL_VALUE=<laburl>
- VITE_GEAI_WORKSPACE_URL_VALUE=<workspaceurl>
- VITE_GEAI_CONSOLE_URL_VALUE=<consoleurl>
- geai-aistudio-frontend container now responds in port 8080.
- If not changed, health check will fail
- Azure Environments
- It is mandatory to Add the following EnvVariable to the API and Console containers GEAI_JVM_MAXRAM=0
- Review and validate the Requests Summary Table.