ignio AI.WorkloadManagement

It is an AI/ML–driven application for proactive workload-management operations. It predicts and prevents batch outages using the combined power of AI/ML and automation to ensure predictable, agile, and silent workload operations. 

It predicts future batch and batch-induced outages, diagnoses root causes, and recommends corrective actions to prevent outages. It leverages the combined power of AI/ML and automation to ensure predictable, agile, and stable workload operations. 

Batch operations teams, batch design and architect teams, and business heads.

The product uses job schedules, dependencies, and historical data to analyze batch behavior and use it to predict a batch, provide ahead-of-time notifications, plan for change, and give recommendations for proactive problem management. Starting with Release 3.8.3, the product’s scope can be used for any process such as orders, transactions, and requests. 

Value benefits can be divided into two buckets: 

Business teams:

• Improved assurance of timely batch runs. 
• Reduced reputation risks by batch-induced outages.

Operation teams:

• Reduced SLA violations.
• Reduced time to resolve job failures. 
• Reduced time to plan changes. 

Following are the key differentiators of the product: 

• Goes beyond predictions to diagnosis and prescription 
• Supports prediction across heterogeneous schedulers 
• Support complex what-if analysis 
• Provides closed-up intelligence by diagnosing and resolving job failures
• Supports historical analysis for proactive problem management

ignioAI.Workload Management uses AI to support four levels of intelligence: 

• Descriptive: Uses unsupervised learning techniques for clustering, classification, and pattern mining to profile normal behavior, mine anomalies, and report batch jobs that need attention. 
• Diagnostics: Uses correlation and causality analysis techniques along with rule-based reasoning to diagnose the cause of batch SLA violations. 
• Predictive: Uses forecasting and regression models to predict future batch 
• Prescriptive: Uses multi-objective optimization techniques along with reinforcement learning to recommend actions to prevent SLA violation.