Anomaly Detector Client
VerifiedDetects anomalies in time series data via stateless or stateful modes.
What is Anomaly Detector Client?
The Anomaly Detector Client exposes Azure's time-series anomaly detection endpoints. It accepts univariate series and returns anomaly scores, boundaries, and change-point indicators according to the chosen detection mode.
It is used by applications that monitor metrics, logs, or sensor readings and need to surface outliers or trend changes without maintaining local models.
Endpoints
3 operations exposed by this API (v1.0).
/timeseries/changePoint/detect— Detect change point for the entire series/timeseries/entire/detect— Detect anomalies for the entire series in batch./timeseries/last/detect— Detect anomaly status of the latest point in time series.What you can build with Anomaly Detector Client
Monitor infrastructure metrics
Developers feed server CPU, memory, and request rates into Entire Detect to automatically flag performance incidents across historical windows.
Detect retail sales anomalies
Retail teams run Last Detect on daily transaction volumes to surface sudden drops or spikes that may indicate stockouts or promotions.
Track sensor drift in IoT
Stateful mode stores ongoing temperature or vibration series so ChangePoint Detect can be called on any recent range without resending full history.
Get started with Anomaly Detector Client
- 1Create an Azure account and provision an Anomaly Detector resource in the portal
- 2Copy the endpoint URL and API key from the resource overview
- 3Install the official SDK for your language (Python, Java, etc.)
- 4Format your time series as JSON arrays of timestamps and values
- 5POST the data to the chosen detect endpoint using the key for authentication
Anomaly Detector Client: pros & cons
Pros
- +Offers both stateless and stateful operating modes
- +Includes change-point detection in addition to point anomalies
- +Simple HTTPS REST interface with official client libraries
- +Integrates directly with other Azure data services
Cons
- –Requires an Azure subscription and billing setup
- –Stateful storage adds ongoing Azure costs
- –Designed for univariate series only
Frequently asked questions
It is an Azure service that finds anomalies and change points in time series data using either stateless or stateful calls.
User reviews
Verified reviews from the community shape this listing's rating.
Loading reviews…