One incident,from the first signalto verified recovery.
This page follows a sample incident: a payment provider starts failing during checkout. Every screen below is the product, with simulated data.
Detect what users feel, not just what servers report.
MonitOps learns the normal shape of each service and journey and flags departures from it. A feed that goes silent is raised as its own problem, so missing data never looks healthy.
checkout-web error rate
Last 30 min · vs 7-day baselineMajor14:02:11 ICT · 6.8 % against a baseline of 0.2–0.7 % · incident opened
Source silent
Unknowncrm-synclogs- Last record
- 14 min ago 13:48:02
- Expected
- every 30 s
Raised as its own problem. crm-sync is shown as unknown, not healthy, until records arrive.
One incident, not two hundred alerts.
Alerts are grouped by time, topology and shared impact, and each incident says why. Alerts that stayed separate show the reason, and an operator can add them.
Correlated alerts in the sample console
Card checkout failing at payment authorisation
- Open
- Investigating
- Mitigating
- Resolved
Correlated alerts · 7 raw → 1 incident
Grouped by shared path in the dependency graph, 14:01:58–14:02:31 ICTpayments-gw5xx responses > 5 %Criticalorder-apiUpstream timeoutsMajororder-apiError rate above baselineMajororder-apip95 latency above baselineMajorcheckout-webError rate above baselineMajorpayments-gwAvailability probe failingCriticalcheckout-webCheckout journey failingMajorcrm-syncLatency· not on the shared pathService context that knows where it came from.
Every relationship says whether it was observed in traffic or declared, and when or by whom. Stale edges fade instead of pretending to be current.
Card checkout · dependencies
INC-20931 path in bold · select a service- Incident path
- Observed from traffic
- Declared
- Stale
payments-gw
External provider gateway · probable cause #1- Health
- Down · 5xx on 31 % of calls since 14:01:58
- Cause
- Provider · relative confidence 0.81
- Owner
- Payments integration
- Journeys
- Card checkout
- Provider
- Status page: degraded card authorisation since 13:59
- Edge
- From order-api · observed, last seen 14:01
Evidence first. Then a conclusion you can check.
Causes are ranked in code from named features, then the AI investigator confirms or refutes each one with cited evidence. Confidence is relative, never a probability. When evidence is thin, MonitOps says so.
Root cause analysis in the sample console
Card checkout failing at payment authorisation
- Open
- Investigating
- Mitigating
- Resolved
Root cause analysis
Ranked candidates · published 14:03:05 ICTpayments-gwproviderpayments-gw 5xx responses 31 % (baseline under 0.5 %) since 14:01:58order-api: upstream timeout calling payments-gw (8000ms) ×1,284 in 4 min13:59order-api healthy on every other dependencyCard authorisation errors start at payments-gw and match the provider's own status report . Every other order-api dependency is healthy .
order-apiintegrationorder-api in 24 hErrors surface in order-api, but it had no changes in 24 hours and its resources are within baseline .
Relative confidence ranks candidates against each other. It is not a probability.
Root cause analysis
INC-20944Insufficient evidence to rank a cause.
No candidate is shown until the feed returns or an operator adds evidence.
Root cause analysis
Computed · scorerInvestigator unavailable. Showing the computed ranking only.
Candidates and their features are unchanged. The AI confirmation step and its citations are missing, and the panel says so.
Thousands of log lines, read as a handful of patterns.
Logs in the incident window are grouped into templates and summarised with citations back to the lines.
- L11,284 lines · first 14:01:5814:01:58
upstream timeout calling payments-gw (<*>ms) - L2212 lines · first 14:02:0314:02:03
order <*> payment state=PENDING retry=<*> - L39 lines · first 14:02:4014:02:40
circuit open: payments-gw
order-api started timing out on calls to payments-gw at 14:01:58. Orders were held as PENDING and retried, and the circuit breaker opened at 14:02:40. No other template rose above its baseline.
Ask in Thai or English. Get answers with citations.
The copilot investigates with the asker's own tools and permissions, and cites every claim.
Copilot in the sample console
Card checkout failing at payment authorisation
- Open
- Investigating
- Mitigating
- Resolved
Root cause analysis
Ranked candidates · published 14:03:05 ICTpayments-gwproviderpayments-gw 5xx responses 31 % (baseline under 0.5 %) since 14:01:58order-api: upstream timeout calling payments-gw (8000ms) ×1,284 in 4 min13:59order-api healthy on every other dependencyCard authorisation errors start at payments-gw and match the provider's own status report . Every other order-api dependency is healthy .
order-apiintegrationorder-api in 24 hErrors surface in order-api, but it had no changes in 24 hours and its resources are within baseline .
Relative confidence ranks candidates against each other. It is not a probability.
Copilot · same incident, different role
viewerShow me the raw payment logs for card ending 4417.
You don't have access to raw payment logs. I can show the masked error summary for payments-gw instead.
See limits before you hit them.
Forecasts carry an explicit uncertainty band and raise a predicted-breach alert while there is still time to act.
sms-gateway API calls
% of daily limit · now 14:20 ICTAlerts
sms-gatewayRestore service, then prove it.
Actions come from a catalogue you control. An operator requests, an approver approves, and the incident resolves only when fresh probes meet the recovery criteria.
Try acting on cause #2 from the step before: the command succeeds and verification still fails.
Recovery in the sample console
Card checkout failing at payment authorisation
- Open
- Investigating
- Mitigating
- Resolved
Recommended action
MonitOps proposes · the policy decidesThe fault is on the provider side and the secondary provider is healthy. Failing over restores card checkout without changing order-api.
provider-failoverEnforced · v3- Mode
- Propose · never runs automatically
- Approvals
- 1 required
- Verifies
- Checkout probes pass and order-api error rate returns to baseline
Timeline
All times ICT- 14:01:58MonitOpsFirst alert: payments-gw 5xx above 5 %
- 14:02:11MonitOpsIncident opened · P1
- 14:02:40MonitOpsGrouped 7 alerts into INC-20931
- 14:03:05MonitOpsRoot cause analysis published
- 14:04:20Narin S.Asked copilot
INC-20931 is resolved. Outside of incidents, teams use the same data for these.
Dashboards your team arranges, including ones you describe.
Arrange widgets on a grid, or describe one and MonitOps builds a validated widget you can keep.
order-api error rate
14:02–14:20New widget
Card checkout path
14:20 ICTcheckout-webHealthyorder-apiHealthypayments-gwDegradedpayments-gw-secondaryHealthycrm-syncUnknown
Reports on a schedule, in the format your stakeholders read.
Incidents, availability and alert volumes on a schedule. Summaries outlive raw telemetry retention, so quarterly figures still work.
- Formats
- PDF · CSV · XLSX · JSON
- Delivery
- Email · webhook · API
- Period
- Any range; recurring daily, weekly or monthly
See MonitOps on your own incidents.
We run a working session with your operations team on your services and recent incidents.