Why per-minute billing for Custom Work?

Custom Work covers automations you direct: scheduled monitors, custom agents, slash commands. The cost depends on how long the agent actually runs, so you pay for the time it works, not for it being available.

How does the free trial work?

14 days on your own production, 500 credits. Cleric joins your existing on-call rotation. Day 12 is a trial review with your team; by day 14 your team has the data to decide.

What happens at the end of the annual contract?

The contract is 12 months and auto-renews. Renewal mechanics, notice periods, and any true-up are covered in the Order Form.

ClericAutonomous AI SRE that triages production alerts and surfaces root causes

4.8 (6)

Reviewed by Daniel Nikulshyn·Updated July 2026

DevOps SRE Observability Incident Response Cloud Native Monitoring AI Agent

Overview

Cleric is an AI-powered site reliability engineer that automatically investigates production alerts the moment they fire. It connects to your existing observability stack, correlates signals across logs, metrics, and traces, and works through hypotheses the way a human on-call engineer would. Instead of waking up engineers for every page, Cleric performs the initial triage, narrows down likely root causes, and delivers a concise summary with supporting evidence. Teams can review its reasoning, confirm findings, and act faster, reducing mean time to resolution and on-call fatigue. Designed for engineering teams running complex cloud-native systems, Cleric aims to handle the repetitive investigative work of incident response so SREs and developers can focus on building and fixing rather than digging through dashboards.

Key features

Autonomous alert triage
Root cause hypothesis generation
Integration with observability platforms
Cross-signal correlation across logs, metrics, traces
Human-readable incident summaries
Continuous learning from production environment

Pricing

Model: Freemium
Category: AI Agents
Rating: 4.8 / 5 (6)

Use cases

Automated On-Call Alert Triage

When production alerts fire, Cleric autonomously investigates and surfaces likely root causes, reducing the need to wake engineers for every page.

Cross-Signal Root Cause Analysis

Correlates logs, metrics, and traces across your observability stack to generate root cause hypotheses backed by supporting evidence.

Reducing Mean Time to Resolution

Delivers concise, human-readable incident summaries so engineers can confirm findings quickly and act, shortening MTTR on complex cloud-native systems.

Easing On-Call Fatigue

Handles repetitive initial investigation work for engineering teams, freeing on-call staff to focus on critical incidents that need human judgment.

Pros & Cons

Pros

Reduces on-call burden by automating initial triage
Investigates across logs, metrics, and traces
Provides reasoning and evidence for findings
Speeds up mean time to resolution
Integrates with existing observability tools

Cons

Effectiveness depends on quality of telemetry data
Still requires human review for critical incidents
Limited value for teams without mature monitoring
May need tuning to fit unique environments

Reviews

4.8

Average from 6 ratings.

Joanna Kowalski

Jan 30, 2026

Solid for our team

We rolled this out across the team last quarter and reduces on-call burden by automating initial triage. Autonomous alert triage fits neatly into how we already work, and integration with observability platforms removed a step we used to do by hand. but it has held up under daily use.

Pierre Dubois

Sep 20, 2025

Compared a few options

Evaluated this against two competitors. Where it wins: root cause hypothesis generation and integrates with existing observability tools. On balance the feature set — especially continuous learning from production environment — justifies the 5 stars for our use case.

Kwame Mensah

Sep 6, 2025

Compared a few options

Evaluated this against two competitors. Where it wins: integration with observability platforms and provides reasoning and evidence for findings. On balance the feature set — especially continuous learning from production environment — justifies the 5 stars for our use case.

Omar Haddad

Jul 29, 2025

Years in this space

I've evaluated a lot of these over the years. What stands out here is integration with observability platforms — handled better than most — and reduces on-call burden by automating initial triage. Still requires human review for critical incidents is my one real gripe. Worth the time if this is your use case.

Tomáš Novák

Jun 25, 2025

Does the job

Pretty happy overall. Continuous learning from production environment just works and integrates with existing observability tools. but no dealbreakers — I'd recommend it to a friend without hesitating.

Aaliyah Johnson

Jun 8, 2025

Compared a few options

Evaluated this against two competitors. Where it wins: human-readable incident summaries and investigates across logs, metrics, and traces. On balance the feature set — especially cross-signal correlation across logs, metrics, traces — justifies the 5 stars for our use case.

Q&A

What does Cleric do for free?

Three categories. Knowledge questions, like “what services depend on the orders API?” System management, like muting an alert or updating Cleric’s memory. Handoffs, like creating a Linear ticket, opening a PR, or posting a Slack message. If it isn’t operational work, we don’t bill for it.

Asked by Quoc Bui · Mar 8, 2026

What counts as an Issue?

An Issue is a defined operational problem Cleric takes end to end. A page from PagerDuty. An anomaly detected by a scheduled monitor. A failed deploy. Cleric investigates, finds root cause, verifies the finding, and proposes a fix. One Issue, one price, no matter how many passes the agent needs.

Asked by Kirsi Laine · Mar 8, 2026

What if Cleric gets it wrong?

Push back and Cleric keeps working the same Issue. We take the risk on getting it right, so a second pass doesn’t cost you more.

Asked by Wei Chen · Feb 18, 2026

What happens if I run out of credits?

Credits roll over one month, capped at 2× your monthly allocation. If you exceed your annual bucket, you can buy more credits, set a hard cap to block new work, or upgrade. Overages are billed on the next invoice and require written sign-off.

Asked by Hannah Goldberg · Jan 30, 2026

Do you train on our data?

No. Customer data is never used to train or improve underlying models. After an evaluation ends, if you don’t continue, we contractually commit to deleting your data. Full security and privacy policy at docs.cleric.ai/security.

Asked by Amara Chukwu · Jan 7, 2026