AI prepares. A person authorizes.
This is not a setting you turn on later. It is the rule the whole system is built on, and the reason Stack Mind can do real work without asking you to gamble your business on it.
Stack Mind rests on one rule: the AI prepares work (looks things up, drafts messages, creates tasks) and a person approves anything that cannot be undone, like sending messages to customers, deleting information or moving money. Each client has their own installation, on their own computer or server, so their information is never mixed with anyone else’s. Each person on the team sees only what their role allows. We do not claim the system can never fail and we do not claim certifications we have not earned. The technical details are explained in the meeting.
Why this rule exists
The most common mistake with AI is the opposite of what we do: give it free rein over a company’s systems and hope it behaves. It works fine in a demo. It fails the first time the AI misreads a situation and what it did cannot be undone.
Stack Mind starts from the other side. Hanna, who does the day-to-day work, and BOSS, who answers from what your company knows, operate with a short list of things they are allowed to do. They can look things up, draft, organize and tell people things. What they cannot do is act in a way that cannot be reversed without a person saying yes first.
This is not caution for its own sake. It is what makes it safe to let an AI actually work in your business instead of just chatting about it. An assistant that only answers questions is easy to trust and useless for running a business. One with no limits is powerful and impossible to trust. Stack Mind lives in between.
The same limit applies everywhere in the product: what the assistant is allowed to do, what information it can see, and which actions stop and wait for you before they happen.
What the AI does alone, and what it never does alone
The line is drawn by what can be undone.
The AI does on its own
- Look up real information in your business to answer a question
- Prepare a draft (a message, a reply, a summary) for someone to review
- Create and assign a task with an owner and a due date
- Tell the right person on the team that something needs attention
- Organize information so it can be found later
- Follow up on something already agreed, without inventing new commitments
- Send a mass message to customers or contacts
- Delete information from your records
- Move money or commit funds
- Change what a person on the team can access
- Publish anything outside: a post, a page, a public message
- Do anything without a record of who approved it
How this is actually enforced
Six controls, in plain language.
A short list of allowed actions
The assistant is given a defined list of what it can do. There is no "everything is allowed" that someone has to remember to restrict.
Passwords stay out of the AI’s reach
Access codes and passwords are handled by the system around the AI. The AI never sees them.
Your own installation
It runs on your computer or your server. Your information is not mixed with that of any other business.
Each person sees their part
What a person can see and do depends on their role in your team.
A record of what mattered
Important actions are logged, so "who approved this and when" has an answer instead of a guess.
What the AI costs, in view
A panel shows what the AI spends, so cost is never a surprise on a bill.
The questions every owner asks
Straight answers, including where the honest answer is "it depends."
In your own installation, on your computer or your server. It is separate from every other business and not scattered across individual phones or inboxes.
The people in your team, according to the role they have been given, plus the assistant acting within the limits you set. Nobody outside your business is part of how it is built.
No. Your business information is used to answer your questions and do your work. It is not pooled with other customers’ information.
The information is yours. If you decide to stop using Stack Mind, we work with you to get it out in a usable form.
We do not claim errors are impossible. No honest company can. What we control is how far a mistake can go: anything that cannot be undone needs your approval first, so a mistake stays a draft, not a consequence.
What we do not claim
- We do not claim a system can never fail. Nobody who is honest with you can say that about any system.
- We do not claim formal security certifications we have not obtained. If that changes, we will say so here.
- Every business is different. In the meeting we tell you plainly what we would set up for yours and what we would leave out.
Ask us anything about how this works
Privacy questions deserve a real conversation, not a page. Tell us what you need to be sure about and we will go through it with your case in mind.