Cleric
Cleric builds an AI site reliability engineer that investigates production regressions and proposes root-cause fixes.
Founded
2023Cleric builds an AI site reliability engineer that investigates production regressions and proposes root-cause fixes.
Cleric keeps every verified investigation in a decision model that later investigations reuse, so the agent accumulates environment-specific knowledge instead of restarting each time. That design targets the main objection to autonomous diagnosis, which is repeated wrong answers in an unfamiliar system.
Read-only default access lowers the security review burden for a first deployment, and the 2025 Gartner Cool Vendor designation in AI for site reliability engineering and observability supplies third-party validation that younger competitors lack.
Source: cleric.ai
Cleric charges through metered credits rather than per-seat licensing, which ties spending to investigation volume instead of headcount. That suits buyers whose incident load grows faster than their engineering teams do.
The model also exposes the vendor to usage variance, since a quiet quarter reduces revenue from the same customer. Enterprise buyers comparing this against the flat platform subscriptions charged by observability incumbents must forecast investigation volume before they can compare cost.
Source: cleric.ai
Cleric maps a customer's services and dependencies, then follows each pull request from open through deployment and watches the released change under real traffic. An alert, regression, support ticket, or engineer question starts an investigation that tests hypotheses against logs, metrics, traces, and earlier incidents.
The agent returns a root-cause explanation with a proposed fix and can pass its production context to a coding agent. Access is read-only by default, and the company reports SOC 2 Type II compliance.
Source: cleric.ai