How you and the AI work together
There's one simple deal behind everything your AI Agent does: the AI proposes, you decide. Nothing about your strategy ever happens behind your back.
The deal in one line
If it would change your strategy or send something out, it waits for you. If it's just gathering facts, it runs.
That's the whole rule. Keep it in mind and you'll always know what to expect.
What runs automatically vs. what waits for you
| Runs automatically | Waits for your approval |
|---|---|
| Gathering research on companies and people | Changes the AI suggests to a contact or company |
| Writing a qualification verdict onto a record | Changes to your targeting |
| Continuing a step you started | Every outreach message, before it's sent |
| Importing your data |
The things that run on their own are all about collecting facts. The moment the AI wants to change something that matters (your targeting, a contact's details) or send a message, it stops and asks you first.
Your edits vs. the AI's
This is the easiest way to feel the difference day to day:
- Your own edits apply instantly. When you change something, it's done. No waiting, no approval step.
- The AI's proposed changes go through Tasks. Each one lands in your list as a simple yes-or-no. You approve it, and it applies. You decline it, and nothing changes.
See Tasks and approvals for how to review and clear these quickly.
Why you can trust it
You should never have to take the AI's word for anything. So we built in three promises:
- Every verdict has a reason. When the AI reaches a verdict, you can read exactly why.
- Every fact has a source. Click through and see where it came from.
- Every change is on the timeline. There's a clear record of what happened and when.
Skim the reason behind a verdict before you act on it. It's the fastest way to build a feel for how your AI Agent thinks.
The AI never sends a message or commits a strategic change on its own. That's always your call.