What your agents can do
- Plan and run marketing campaigns, and write and design the creative
- Answer your inbox — email and SMS — and follow up with leads and customers
- Draft and manage content across the company site and channels
- Keep your integrations wired up and healthy
- Track progress against your goal and re-plan each cycle
What they can't do
The guardrails aren't a setting you turn on — they're how the system is built.
- Spend money, email customers, or publish public changes without clearing the guardrails (and, where it matters, your approval)
- Take actions outside the catalog of things Reeve is built to do — off-catalog requests are refused, not improvised
- Invent your numbers — goal totals and metrics are computed server-side from real events, not estimated by the model
- Move you toward ownership or bill you in ways that differ from what's on the pricing page
Human in the loop
Every consequential action routes through an approval queue before it happens. You approve it, edit it first, or reject it — and the agents take that signal into their next cycle. You choose how much runs unattended; nothing that spends money or reaches a customer bypasses that choice.
Common questions
How much control do I actually have?
What stops an agent from going rogue?
What does “operated by AI” really mean here?
How is my data handled?
Can I get a refund?
What happens if I stop paying?
Where to go next
See how approvals work in the cockpit & your C-suite, and how billing and ownership work in how ownership works.
Still need a hand?
Ask your Concierge right inside the cockpit, or email support@meetreeve.com and a human will get back to you.