Squadbase Member Activity
How to use the Squadbase Member Activity connector, which turns your team's AI chat conversations into a data source so you can analyze how AI is used
Squadbase Member Activity is a connector for analyzing how your team uses AI. Connect your organization's AI chat conversations as a data source, and just ask AI questions like "Which projects use AI the most?" or "What kinds of questions come up most often?"
You can use what you learn about AI adoption to train your team, improve your knowledge, and explain the impact of rolling out AI.
Squadbase Member Activity is available to admins of organizations on the Team plan or higher. For organizations below the Team plan and for members who aren't admins, it is shown as locked, along with the reason. → Plans
What you can analyze
- Which projects use AI the most
- What kinds of questions are most common
- How each member uses AI
You can analyze it just by asking AI in a project's chat, without writing SQL. You can also analyze it with SQL.
To analyze AI credit consumption, use Squadbase Credit Usage. Squadbase Member Activity is for analyzing the content of conversations; Squadbase Credit Usage is for analyzing credit consumption.
As soon as it's created, you can start analyzing from the most recent conversations. Past conversations are made available for analysis in the background, one after another.
Connection details
The connection details show whether the data is up to date with the latest conversations and the period available for analysis. Admins can click Sync now to include the latest conversations in the analysis right away.
Period available for analysis
| Plan | Period available for analysis |
|---|---|
| Team | The last 12 months |
| Enterprise | Set individually per contract |
Conversations older than this period are automatically excluded from analysis. Even when excluded from analysis, the original conversations are not deleted.
If you change your plan to one below Team, syncing stops automatically at the next sync. If you go back to Team or higher, it resumes automatically.
How data is handled
- Conversations are prepared in a dedicated analytics database for each team
- Who said what is handled not by user ID or email address but by a pseudonymous ID and display name that are only meaningful within that team
- Secrets in common formats, such as tokens and private keys, contained in conversation text are masked before being added to the analytics data
- Conversation search uses partial matching. You can also search for Japanese text
Related pages
- Audit log — A history of important actions in your organization (Squadbase Audit Log)
- Connectors