Overview
Cloud costs can fluctuate for many reasons — scaling, configuration errors, or unexpected workloads. Without monitoring, these changes may remain unnoticed until they result in unnecessary overspending. Uniskai's Anomaly Detection feature empowers users to set flexible thresholds, automatically track consumption, and receive alerts whenever usage crosses defined limits.
Notifications are sent by email, and — if Slack is connected — to your personal Slack and to a channel via webhook as well. Read more about notifications here.The Anomaly settings are located in the Organization Settings, please refer to this manual for access instructions.
To confirm before publishing: that anomaly notifications support the channel webhook; if not, drop the webhook mention.
Where to find the settings
The Anomaly Detection settings are located in Cloudcosts → Cost monitoring → Anomalies. Open the Cost monitoring tab from the left navigation and select the Anomalies sub-tab.
Note: Only organization owners or admins can edit anomaly settings.
How thresholds work
Anomalies are identified based on configured thresholds. Owners or Admins can define both a percentage threshold and a cost threshold, and then choose whether these conditions should be applied together or separately:
AND — anomalies are triggered only when both the percentage and cost thresholds are exceeded.
OR — anomalies are triggered when at least one of the conditions is met.
This flexibility allows teams to fine-tune detection rules to match business needs.
Daily and monthly monitoring
Uniskai supports two monitoring schedules: daily and monthly. Each can be configured independently with its own thresholds and notification settings. Daily monitoring is useful for detecting short-term spikes, while monthly monitoring highlights longer-term consumption trends.
Both daily and monthly monitoring can be completely turned on or off at any time, giving organizations full control over how anomalies are tracked.
Global and account-level thresholds
Thresholds can be applied globally or at the account level:
Default thresholds cover all accounts unless a specific account threshold is defined.
Account-level thresholds allow specific accounts to be monitored differently from the rest.
If an account does not have an account-level threshold, the system automatically applies the default values.
In addition, anomaly detection can be managed per account. Users can disable anomaly searches for specific accounts directly from the interface or update thresholds for individual accounts when needed. Available actions:
Update — set custom thresholds for the account.
Default — reset the account back to the default (global) thresholds.
Disable — turn off anomaly detection entirely for the account.
Viewing detected anomalies
On the Anomalies tab you can review detected anomalies and the underlying spend:
Total billing — a chart of your spend across the selected accounts.
History — a table of detected anomalies showing Anomaly source, Account/ID, Previous day's consumption, Expected consumption, Actual consumption, Detected anomaly (amount and %), and Date.
Show in Graph — from any History row, open the anomaly on the chart to see the affected service highlighted on its date.
Notifications
When an anomaly is detected, alerts are delivered through the configured channels:
Email — sent to the notification recipients configured in the anomaly settings.
Slack — if connected, delivered to your personal Slack and/or a channel via webhook.
To confirm before publishing: exact content of the anomaly email/Slack message and where notification recipients are configured for anomalies.
Summary
By combining threshold configuration, independent daily and monthly monitoring, and per-account controls, Uniskai's Anomaly Detection feature ensures that no unusual consumption goes unnoticed. This proactive approach helps organizations prevent overspending, identify misconfigurations, and maintain effective cost management.