Question 1

Is it actually safe to ask an AI about my clients?

Accepted Answer

Yes, and this is the whole reason Ask Clerie exists. It runs inside Clerie, under the same HIPAA compliance as the rest of your chart, and your client data is never used to train AI. Every login is bound by role-based permissions, so the assistant only answers with what that person is cleared to see. Compare that to the workaround it replaces: pasting client details into ChatGPT, which is a HIPAA violation waiting to happen. ChatGPT can’t legally sit this close to a chart. Ask Clerie was built to.

Question 2

Can it change things in my clinic without me knowing?

Accepted Answer

No. Answers are instant; actions wait for you. When you tell Ask Clerie to do something, like putting John on the calendar every Tuesday at 3pm, it shows you exactly what it’s about to do and nothing happens until you say yes. There’s no background automation quietly rescheduling your week. And because every action runs through your clinic’s permission system, a login that isn’t allowed to change the schedule can’t get Ask Clerie to change it either. The confirm step is the design, not a setting you have to find.

Question 3

Will we hit a token limit if the whole staff uses it?

Accepted Answer

No practical limit. We’ve never had a clinic run out, and we’re not going to nickel-and-dime you per question. Ask Clerie is included with Clerie: there’s no per-seat AI fee and no usage meter your office manager has to watch. That matters because the value shows up when everyone uses it, from the front desk asking how to resend an intake form to the owner pulling a month-end export. If the whole team asks it questions all day, that’s the tool working, not a bill building.

Question 4

What can the front desk see if they ask it?

Accepted Answer

Only what the front desk is cleared to see. Ask Clerie inherits your clinic’s role-based permissions, the same ones that govern every screen in Clerie. If a login can’t open a chart, it can’t ask its way into one either. When the front desk asks a question, the answer draws on front-desk data. When the owner asks, it can pull the whole picture. You set the roles once and the assistant respects them on every question. There’s no separate AI permission system to configure, and no way to ask around the rules.

Question 5

What happens when it gets an answer wrong?

Accepted Answer

You thumbs-down it, and Clerie learns from it. Every piece of feedback goes back into making the assistant better at your clinic’s questions, so the longer your team uses it, the sharper it gets on the things your team actually asks. And a wrong answer can’t turn into a wrong action: anything that would change your clinic still shows you a confirmation first, so a miss costs you a thumbs-down, not a double-booked Tuesday. If it whiffs on something important, tell us. Training it is part of the product, not a favor to us.

Question 6

What can Ask Clerie not do?

Accepted Answer

It works inside Clerie, on Clerie. It answers questions about your clinic’s data and the product, takes actions you confirm, and builds exports. It is not a clinical advisor: it will not diagnose, recommend treatment, or write anything into a client’s official chart on its own. If you ask for something outside its lane, it says so instead of guessing. And anything that changes your clinic still shows you a confirmation first. That boundary is deliberate. An assistant you can trust is one that knows what it’s for.

ChatGPT for your clinic. That takes action.

Every “how do I” lands on you.

Ask a question, or tell it what to do.

Ask a question, get the steps

An AI agent that takes action

Excel exports without the report builder

Your permissions and HIPAA still apply

Answers, actions, and exports in one chat box.

As powerful as ChatGPT, and HIPAA-compliant.

The questions people actually ask on demos.

Ask it something on your demo

Product

Who it's for

Plans

Company

Legal & support

ChatGPT for your clinic. That takes action.

Every “how do I” lands on you.

Ask a question, or tell it what to do.

Ask a question, get the steps

An AI agent that takes action

Excel exports without the report builder

Your permissions and HIPAA still apply

Answers, actions, and exports in one chat box.

As powerful as ChatGPT, and HIPAA-compliant.

The questions people actually ask on demos.

Is it actually safe to ask an AI about my clients?−

Can it change things in my clinic without me knowing?+

Will we hit a token limit if the whole staff uses it?+

What can the front desk see if they ask it?+

What happens when it gets an answer wrong?+

What can Ask Clerie not do?+

Ask it something on your demo