TEADAL Unveils AI-Driven Performance Monitoring System to Optimise IT Resource Utilisation

Authored by Almaviva 

TEADAL has introduced a cutting-edge AI-Driven Performance Monitoring (AI-DPM) system aimed at revolutionising how IT systems are monitored and optimised. This new approach leverages AIOps to provide real-time anomaly detection and predictive insights into resource utilisation, particularly focused on enhancing efficiency and system performance.

At the heart of the AI-DPM system is a robust metadata management framework that collects time-stamped runtime metrics via a sophisticated stack of observability tools. Prometheus, Kepler, and Istio each play a key role in this system:

• Prometheus captures system resource performance metrics,
  
• Kepler monitors energy consumption, and
  
• Istio delivers service-level observability.

These metrics are aggregated and stored in Thanos, enabling advanced querying capabilities through PromQL. This setup facilitates the training of machine learning models that predict future resource utilisation trends. For instance, by focusing on CPU usage and defining parameters such as training hours, input steps, and output steps, the system can provide average CPU utilisation per node — a crucial input for both training and forecasting.

The resulting predictive insights, combined with historical data, empower TEADAL to make informed decisions about system optimisation and energy efficiency. Ultimately, the AI-DPM transforms raw monitoring data into actionable intelligence, supporting TEADAL’s mission to advance sustainable and intelligent IT operations.

This demonstration marks a significant step toward AI-powered automation in performance monitoring, positioning TEADAL as a leader in data-driven infrastructure management.