Veritus/Journal/Checklists and swipe files

How vv decides to send, ask or wait: 6 trust rules

An AI that talks to your customers needs rules, not just intelligence. These are the six we are building into vv.

  1. Drafts first. Nothing goes out until you have seen what it sounds like.
  2. Ask before money. Refunds, discounts and payments always need a human yes.
  3. Autonomy is earned. It sends the easy, routine messages on its own only after you have approved them many times.
  4. One question when unsure. It asks you, once, rather than guessing.
  5. Every action leaves a trail. You can see what was sent, when and why.
  6. Quiet hours. It does not message customers at night unless you say it is urgent.

Why this matters

The risk with AI in a small business is not that it is clever. It is that it is confident and wrong. These rules make wrong cheap and visible.

You can use the same rules with any tool, or with a human assistant. See how to implement AI step by step.

Which rule would you add?

The best answers go into vv. Message Abhisar.

Message Abhisar on WhatsApp →

Frequently asked questions

How should an AI agent behave with customers?

It should draft first, ask before anything involving money, earn more autonomy over time, ask one question when unsure, leave a record of every action and respect quiet hours.

Talk to Veritus

Tell us what keeps landing on your desk.

Send one line on WhatsApp about your business and the job that eats your week. Abhisar replies personally within one business day and tells you honestly whether it needs a tool, a process fix or nothing at all.