Skip to main content
Analytics shows how people are using Upsolve: how many conversations they have, what they save, and who is most active. Open it from Monitoring → Analytics in the workspace sidebar. It opens on Usage Analytics, a board of charts over your organization’s usage. Every number on it is for your organization only.
The Usage Analytics board: totals for canvases, chats, shares and users over the last 7 days, and a chart of weekly chats by user

What the board shows

  • Key totals: how many Canvases, Chats, Shares and Users there were in the last 7 days, and the change from the 7 days before.
  • Weekly chats by user: conversations per week, for your top users, with everyone else grouped as Other.
  • Chats by user, Favorites by user and Canvases by creator: who is doing the most of each.
The chat figures count your end users’ conversations, not builders testing in the Playground. The board is view-only. Use + Add filter to narrow it.

Ask about your usage

For anything the board doesn’t show, click Ask AI. Analytics is itself an Upsolve analytics agent over your usage data, so you can ask in plain English, for example “which users asked the most questions last month?”, and get an answer with a chart. Your earlier questions are kept for next time.

How it differs from Chat History

Chat History is a list of one workspace’s conversations, row by row, with their evaluations, feedback and cost. Use it to inspect a particular answer. Analytics is the bigger picture: totals, trends and who is using Upsolve, without adding it up yourself.

Who can see it

  • Builders only. Owners, admins and members can open Analytics. Read-only members and end users can’t.
  • Not in every deployment. Analytics isn’t available in Upsolve Local, or in a self-hosted deployment that hasn’t been set up for it. There the page says “Analytics isn’t available here”.
If the page says “Analytics couldn’t be loaded”, click Retry.

Next steps

Chat History

Inspect individual conversations, with their evals, feedback and cost.

Canvas

Build boards like this one over your own data.