How to use the AI Analytics screen (I8I Intelligence)
Assistant volume and quality for the last 30 days. Token usage and cost are not recorded anywhere, so none is shown.
Have questions or need step-by-step guidance on this topic?
Step-by-Step Procedure
Open the page
Sign in as an administrator and use Admin sidebar → I8I Intelligence → AI Analytics, or go to /admin/i8i/analytics. The heading reads "I8I AI Analytics".
Know where the data comes from
This page reads i8i_chat_messages and i8i_support_requests via the chat analytics. KPI tiles Turns, "User questions", "Guest share", "Mean latency", "Resolution rate", "Unanswered rate", "Escalation rate", "Support requests". A table by topic, language, mode and tool.
Read the rates
Rates read "no data yet" before any turn exists. For super-admin chat metrics with a date window see Chat Analytics (/admin/i8i-analytics).
Use the shared controls
Use the shared controls: the search box ("Search records, hashes, DIDs...") filters rows by any text; "Status:" offers ALL, TOPIC, LANGUAGE, MODE, TOOL; "Page Size:" sets 10, 15, 25 or 50 rows; Refresh reloads; Export saves the rows currently loaded as a JSON file. The line "Data:" under the heading names the real source.
Computed assistant metrics.
Troubleshooting & Failure Modes
Was this guide helpful?
Help us refine our platform documentation