Custom Analytics Concepts and Lifecycle
Custom Analytics are user-defined data analysis functions that can be created, managed, and shared across companies within a domain. They allow organizations to implement specialized data processing workflows tailored to their specific needs.
Prerequisites
The Custom Analytics operates within a hierarchical structure:
- Domain: Contains multiple companies
- Companies: Organizations that can create or use analytics
User roles :
- Developer: Can create credentials, develop custom analytics, and manage their availability
- Manager: Can enable custom analytics for use within companies
- User: Can execute (launch) custom analytics on projects
Custom analytic LifecycleWorkflow
Development Phase
The lifecycle of a Custom Analytic follows a clear progression from creation through sharing and eventual retirement:
| Steps | Who does it | Description | When to do it |
| 1- Configure credentials | developer | Establish access to required resources and Docker registries | Before custom analytic creation |
| 2- Create Custom analytic | developer |
Create the custom analytic in your company After this step, the custom analytic is enabled and available only in the developer's company |
For each creation or new version updates |
| 3- Launch custom analytic | developer | Verify functionality before sharing by launching the analytic in a project within the developer's company | After the creation to test the custom analytic |
The CLI command are detailed in this article
Sharing Phase
After testing, to make the custom analytic available to other companies:
| Steps | Who does it | Description | When to do it |
| 4- Expose Custom analytic | developer | Make analytic available in your domain. The analytic becomes available in the domain but is not yet visible or usable by other companies | When it's ready to share |
| 5- Enable Custom analytic in target companies | developer, Manager |
Two cases :
After this, authorized people in those companies can use the custom analytic |
When it's ready to be used by other companies |
The CLI command are detailed in this article
Maintenance Phase
After the custom analytic is in use, it may need to be managed:
| Steps | Who does it | Description | When to do it |
| Update credentials | developer |
Update the credential information with the following CLI command alteia analytics set-docker-credentials-name <analyticname> --version <version> --company <short name of the company> --docker-credentials-name <name of the credentials to use to pull the dockerimage from the registry> |
When access requirements change |
| Unexpose Custom analytic | developer |
Making an analytic unavailable across your domain when you no longer want it to be launched in other companies. To unexpose an analytic from a domain: alteia analytics unexpose <analyticname> --domain <domain_name> |
When you no longer want to share it |
| Disable Custom analytic | developer, manager |
Disabling an analytic in your companies, preventing it from being launched To disable an analytic across all domain companies: alteia analytics disable <analyticname> --domain <domainname> To disable an analytic for specific companies using the CLI or administration module alteia analytics disable <analyticname> --company <comma_separated_list_of_company_IDs> |
When the custom analytic should not be used within a company or a domain |
| Delete Custom analytic | developer |
Permanently removing a version of a custom analytic from the platform To delete a version of custom analytic: alteia analytics delete <analyticname> --version <version> |
When the custom analytic is obsolete or no longer needed |
Unexpose custom analytic
If an analytic was previously exposed and activated on some companies, unexposing it will make it invisible to all users. Re-exposing it will automatically make it visible again to users of previously activated companies..
In Simple Terms
Think of Custom Analytics like recipes:
- A chef (developer) creates and tests a recipe
- When perfect, they share the recipe with other restaurants (companies)
- Each restaurant manager decides whether to add it to their menu
- Customers (users) can then order the dish when it's on the menu
- The recipe can be updated, removed from certain menus, or deleted completely if needed
