Metrics

Visualizing the running ML system is important to measure success rate of the system. There is a dashboard which you can import into your grafana installation and can be found here: https://bit.ly/2ppyFO0

Fact Store Metrics

Metrics for seeing the successful deployment of factstore. We may add more metrics in the future.

Metric	Details	Metric Type
aiops_human_feedback_count	count of number of human feedback provided by customer	Counter
aiops_human_feedback_error	count of human feedback not able to write to db	Counter

Core Metrics

Metrics for visualizing the running ML job running and false positives found. We may add more metrics you will find

updates here.

Metric	Details	Metric Type
aiops_lad_train_count	count of training runs	Counter
aiops_lad_inference_count	count of inference runs	Counter
aiops_lad_anomaly_count	count of anomalies runs	Gauge
aiops_lad_anomaly_avg_score	avg anomaly score	Gauge
aiops_lad_false_positive_count	count of false positives processed runs	Counter
aiops_hist	histogram of anomalies runs	Histogram
aiops_lad_threshold	threshold of marker for anomaly	Gauge