Skip to main content
This document provides a comprehensive overview of TRU PULSE, a core component of the Trusys AI Assurance Platform. It focuses on monitoring and managing your AI applications in production, covering monitoring settings, dashboards, traces, and alerts. The Track section allows you to analyze the historical performance and behavior of your monitored applications. Select the application you want to track and the desired date range for analysis.
  1. Application Health Dashboard – This dashboard provides a high-level overview of your application’s operational health:
    • Total Logs Collected – The total volume of logs ingested by Trusys for the selected period.
    • Avg Request Duration – The average time taken for your application to process requests.
    • Avg Token/Request – The average number of tokens processed per request (relevant for LLM-based applications).
    • Avg Cost/Request – The average cost incurred per request, helping you manage operational expenses.
    • Log Request Per Time Series – A visualization of the volume of log requests over time, identifying peak usage or periods of inactivity.
  2. Security Dashboard – This dashboard focuses on the security posture of your application:
    • Total Logs Evaluated – The number of logs that have undergone security evaluation.
    • Total Security Probes Generated – The total number of security tests or probes created, which is a factor of the categories monitored and the logs evaluated.
    • Vulnerabilities Detected – The count of identified security vulnerabilities.
    • Percentage of Security Probes that Failed – The proportion of security tests that revealed a vulnerability.
    • Vulnerability-wise Analysis – A detailed breakdown of vulnerabilities by type, including total probes, pass count, fail count, and severity type.
  3. Functional Dashboard – This dashboard provides insights into the functional performance of your application:
    • Total Logs Evaluated – The number of logs that have been functionally evaluated.
    • Logs Passed, Failed Count – A summary of logs that passed or failed their functional metric evaluations.
    • Metric Failure Analysis – Identifies the metrics that are failing most frequently, with an option to view detailed information for each.
    • Alert Notification Timeline – A chronological timeline indicating when alerts were triggered based on your defined alert settings.
    • Metric-Volume Distribution – A visualization showing the distribution of metric scores, helping you understand the typical performance range and identify outliers.