Integrating AI with Maximo IT

Maximo AI Service

Maximo AI Service is the integration hub that facilitates communication between Maximo IT and watsonx™ AI systems or services. Maximo AI Service facilitates the following aspects of the AI integration:

• Manages configuring, training, and retraining AI models and retains data during training.
• Delegates inferencing jobs to watsonx™ AI or to a local embedded runtime.
• Completes health checks of the AI model runtime and the individual models.
• Integrates into Maximo IT the AI capabilities that enable AI features, such as recommendations.

Maximo AI Service supports multitenancy. Maximo AI Service uses Granite™ 20b Multilingual and Granite 13b models and uses a version of watsonx AI that is hosted in a cloud environment.

For more information, see Deploying Maximo AI Service.

By integrating with the Maximo AI Service, Maximo IT users can unlock machine learning models to automate ticket assignment recommendation and identify possible solutions and similar incidents across historical ticket data—ultimately improving accuracy, boosting user productivity, and reducing response times.

Maximo IT in collaboration with the Maximo AI Service can help organizations leverage two powerful, machine learning models:

• Multi-Class Classification (MCC) for dynamically categorizing incoming requests.
• Similarity Model for identifying related or duplicate incidents from historical data.

What this feature enables

• Smart ticket recommendations.
• Accelerated resolution of issues via historical similarity insights.
• Boost productivity of agents and SRE's through quickly presented solution ideas and similar ticktets to gain insights from.
• Consistent service quality, even across large or distributed teams.
• Continuous learning, with models improving over time as they’re trained with actual incident data.

Who this benefits

• Service Desk Agents can reduce time spent searching for solutions and work going on in similar incidents.
• Site Reliability Engineers (SREs) gain insights into what issues can be addressed through automation and improved code or product use.
• Incident Managers gain visibility into recurring issues and resolution patterns.
• IT Administrators can fine-tune models and monitor performance from a central location.
• End Users see faster turnaround and more relevant solutions through improved automation.

The AI configuration in Maximo IT involves a three-stage setup process:

1. Establishing a Connection with Maximo AI Service - Begin by securely connecting Maximo IT to the Maximo AI Service, which hosts and manages the machine learning models used for classification and similarity detection. This connection enables real-time access to AI models during service operations.
2. Configuring MCC and Similarity Models - Once connected, configure the Multi-Class Classification (MCC) model to recommend incidents to work with, and the similarity model to detect similar tickets and possible solution records. These models can be fine-tuned to align with your organization’s ticket taxonomy and use cases.
3. Training the AI Models - To maximize model effectiveness, train the MCC and similarity models using relevant historical data. This step ensures the models learn from actual usage patterns, improve over time, and deliver context-aware predictions tailored to your environment.