Langtail’s observability features allow you to get an overview of what’s happening to your application both during development and in production. You can look at logs to find out where you app isn’t returning the best result and at metrics to see how latency and cost are affecting your app.

Here’s a quick overview of both:

Logs

Whenever a deployment is invoked or a prompt is sent via the OpenAI Proxy, a log is created. You can view logs in two different places:

  1. When you open a prompt, click the Logs tab in the navigation bar. Here you will see all logs from this prompt.
  2. From the project view, click the Logs tab on the left sidebar. Here you will see logs from every prompt in the project.

From the logs view, you can use filters to drill down by environment, version, prompt, and more. When you click on a log, you get details about how the prompt was invoked. You can also click the Open in Playground button to open it with all of the context from the log.

Metrics

Just like in the logs, you can view metrics in two different places:

  1. When you open a prompt, click the Metrics tab in the navigation bar. Here you will see only the metrics from this prompt.
  2. From the project view, click the Metrics tab on the left sidebar. Here you will see aggregated metrics from every prompt in the project.

You can filter by time in the top right corner of the app.