Integration Components

Edit online

Maximo IT components are used to integrate AI with your business use case and support Maximo AI Service connectivity.

Before you begin

Before configuring AI capabilities in Maximo IT 9.1, ensure the following prerequisites are met. These components are essential to establish connectivity with the Maximo AI Service, manage model deployment, and maintain secure and reliable operations.
  1. Platform Requirements
    • Maximo Application Suite 9.1 must be installed and operational.
    • Maximo IT 9.1 application is deployed with necessary licenses activated.
    • Maximo AI Service is installed and accessible (can be hosted on OpenShift or another supported platform).
  2. Access and User Roles
    • Admin access to Maximo IT and the Maximo AI Service instance.
    • Valid credentials with permissions to perform model configuration and training.
    • Access to MAS Core services such as Application Points, Security Services, and Connectivity.
  3. Data Availability
    • Historical incident and service request data available in Maximo IT for training AI models
    • Data is structured and categorized according to the organization’s service taxonomy (ideally using consistent class structures)

Each available model has an associated template. The following table contains each available template and a list of required integration components for that template.

Table 1. Required components per template
Template Description Required components Other requirements
Multi-class Classification (MCC) Categories incoming service tickets
  • Object structures
  • Object structure saved queries
  • Object structure query templates
  • System properties
  • Endpoints
  • Invocation channels
  • Cron tasks
 Training and inferencing
Similarity Detects historically related or similar incidents
  • Object structures
  • Object structure saved queries
  • Object structure query templates
  • System properties
  • Endpoints
  • Invocation channels
  • Cron tasks
 Training and inferencing
Object structures

They also define the API set that is used to configure the model for the specified Maximo IT business object. Object structures also define permissions for users and groups who are authorized to use Maximo IT AI features.

For more information about object structures, see Object structures.

Object structure saved queries

Osclause queries that are used to select a training data set and control the inference set. The osclause query contains an SQL where clause, without the WHERE keyword, that is based on the root object of the object structure.

For more information about object structure queries, see Object structure query.

Object structure query templates

Define the schema that is used for training and inference requests.

System properties

Configure the URL, the API key, and the tenant ID for Maximo AI Service.

Endpoints

Communicate to external services, such as the Maximo AI Service APIs. The Maximo AI Service API is a Java™ application that enables the flow of data between Maximo Application Suite and watsonx.ai. The invocation channels introduce system property values to endpoint properties for training, inference requests, model status checks, and Maximo AI Service connectivity health checks.

Invocation channels

Define the processing logic and mapping of inbound and outbound data, which enables Maximo IT to process responses to and from Maximo AI Service. They also prepare training and inference data from Maximo IT for the selected object structure and object structure query template.

You can start an invocation channel in the Maximo IT UI or use Cron tasks to start the invocation channel on a schedule.

Cron jobs

A Cron job for scheduling model training and a Cron job for inference requests are available by default.