diff --git a/images/wrangler_insights.png b/images/wrangler_insights.png new file mode 100644 index 00000000..16f44e48 Binary files /dev/null and b/images/wrangler_insights.png differ diff --git a/why-promptlayer/analytics.mdx b/why-promptlayer/analytics.mdx index ac009646..8afcdf0e 100644 --- a/why-promptlayer/analytics.mdx +++ b/why-promptlayer/analytics.mdx @@ -15,6 +15,12 @@ Here you can find key performance indicators to assess your application's perfor Understanding usage patterns is crucial for optimizing your application and improving user experience. Analyzing usage patterns involves exploring prompt registry states, model distributions, tokens and requests over time, latency and cost analytics, and prompt template overall costs. These features provide insights into how prompts, models, and resources are utilized, helping you make informed decisions to enhance your application's performance. +## Ask Wrangler for request-data insights + +If you want a quick read on request usage, open the Wrangler assistant in the bottom-right of the screen and ask directly (for example: "Create a graph of most used prompts over the past week"). Wrangler can probe request analytics first, confirm which prompts have traffic, and then build a chart view from that data. + +![Wrangler insights in Prompt usage analysis](/images/wrangler_insights.png) + ## Filtering and Organization To streamline your analysis, the analytics page offers filtering options based on metadata and tags.