Device Reliability

Device Reliability turns raw OS event telemetry into a single, actionable score for every managed device, making it easy to spot which machines are quietly degrading before they cause downtime.

Composite Scoring Across Five Factors

Each device receives a 0-100 reliability score derived from uptime percentage, crash frequency, application hang rate, service failure count, and hardware error severity. Factors are weighted so that the metrics with the greatest real-world impact on stability carry the most influence.

Trend Direction and MTBF

Breeze computes a trend direction — improving, stable, or degrading — using linear regression over 30 days of daily reliability estimates. Mean Time Between Failures gives teams a quantitative handle on how often a device experiences critical events, supporting hardware refresh prioritization and proactive replacement decisions.

Four Risk Bands for Fast Triage

Scores map to Critical, Poor, Fair, and Good bands. Fleet views can be filtered by band and trend, so operators can immediately surface the devices most in need of attention without scrolling through hundreds of entries. The organization summary highlights the ten worst-scoring devices and counts by band at a glance.

Cross-Platform Event Collection

The Breeze agent collects reliability telemetry from Windows Event Log, macOS system logs, and Linux journald. Crash types, application hangs, service failures, and hardware errors including BSODs, kernel panics, OOM kills, and WHEA events are all tracked natively per platform.

AI-Assisted Fleet Analysis

The Breeze AI assistant can query reliability data using natural language, answering questions like which devices have the worst scores, how many are in a degrading trend, or what the average reliability is for a specific client. This makes fleet-wide reliability analysis accessible without writing queries or building custom reports.