Custom Metrics & Analytics
Learn how to query, create, and update custom metrics programmatically
Last updated
Was this helpful?
Learn how to query, create, and update custom metrics programmatically
Last updated
Was this helpful?
Custom metrics empower you to evaluate and monitor key aspects of your applications by creating tailored metrics that fit your specific needs. Using Arize Query Language (AQL), a SQL-like language, you can easily define and implement these metrics to gain deeper insights into your application performance, cost, usage, and more.
With custom metrics, you have the flexibility to:
Analyze model performance using precision, recall, and other key metrics.
Track costs and usage patterns, helping optimize your application.
Create entirely custom metrics using conditionals, arithmetic, and other custom logic.
Related Pages:
Syntax Guide: Learn the basics of AQL and how to write custom metrics from scratch.
Examples: Explore example metrics to see how you can apply AQL to common use cases.
AQL Generator: Arize's custom metrics, combined with Copilot's intuitive query generation, make it easier than ever to monitor and enhance your AI applications.
