Agents where the work already is
If your team lives in Google Workspace, the data and the permissions are already there. Studio adds the loop: a trigger, a set of steps, and a model that reads and writes your documents.
An email, a form, a schedule
Agents start from things Workspace already knows about: a message arrives, a row is added, Monday morning comes around.
Read, decide, write
Each step reads from a Workspace source, asks Gemini to do something with it, and writes the result back to a doc, sheet, or reply.
Your access, not more
An agent can only see what you can see. That is the safety property, and also why a shared agent may behave differently for different people.
What you will do in the lesson
You answer and try things, step by step.
- Pick a weekly chore. You choose one recurring task, like summarising a shared inbox, and write what a good result looks like.
- Set the trigger. You make the agent run on a schedule and confirm it fires.
- Write the instruction. You tell the model what to read and what to produce, with an example of the output you want.
- Route the output. You send the result to a Doc and to a Chat space and check both.
- Review before trust. You end by reading a week of outputs and deciding what the agent may do alone and what needs a person.
Before you start
Do I need a paid Workspace plan?
Studio is part of Google Workspace with Gemini features, so availability depends on your plan and admin settings.
How is this different from Zapier?
Zapier connects hundreds of apps. Studio stays inside Workspace, which makes it simpler and safer if that is where your work already lives.
Can it send email on my behalf?
If you let it. The lesson's last step is exactly about deciding which actions to allow without review.
Is it underrated?
The Scholé team thinks so. Most people never look past the chat box in Gemini and miss that the same model can run small jobs for them.
