Chiron is an AI companion you configure and supervise for each patient: a steady presence between sessions, grounded in the plan you set, answering in the voice and modality you choose, and accountable to you as the clinician of record.
Cancel anytime, and your records leave with you.
Unsupervised AI forgets your patient the moment the conversation ends, knows nothing of your plan, and is built to keep them talking rather than to help them get better. Chiron is built the other way around. It keeps a lasting memory of each patient, follows the direction you set, carries a hard rule against ever offering a diagnosis, and hands everything it learns back to you. The same reach your patients already trust, now working for their progress and answerable to you.
One companion, facing two ways. To you, Chiron is an invaluable assistant: it carries your plan into the week, keeps watch for what you flag, and turns each patient’s week of chats into a short brief you read before you sit down together. To your patient, it is a shepherd: a steady presence between visits that remembers them and stays inside the bounds of your care.
It does not stand in for you. It extends your reach.
Two views of the same companion. Slide to see the other side.
You’re seeing what the therapist sees. Slide to see your patient’s side.
Chiron is not one companion. It is the one you build for each patient. Choose the modality and the voice. Decide how much it challenges and how much it reassures. Give it the history, so it starts where you already are, not from scratch. Name what it should watch for. And nothing it says comes from nowhere: it all traces back to what you set.
That's a hard thing to carry into a Sunday evening, especially when there's nothing you can do to find out yet.
What does the "it's bad news" thought have to go on so far? Is there anything concrete, or is it mostly the not-knowing?
Your worry window is coming up at 6:00 this eveningTechnique she was taught. If the replaying starts before then, you can tell yourself you'll give it those twenty minutes, and write down what the auditorHer own name for the critical voice is saying and what you actually know.
That's a hard thing to carry into a Sunday night, especially with nothing you can do about it until tomorrow.
Right now, the only solid fact is that your boss wants to talk and didn't say why. Everything past that is your mind filling in the blank, and it's been filling it with the worst version. That's understandable after the arguments with the sous-chef this monthWhat he's working through, but it's still a guess.
Let's check the factsDBT skill he was taught. What do you actually know about where things stand at work? Has anyone said something to you directly, or is this mostly the worry talking?
If it keeps looping, write down what you're predicting, something like "I'm getting fired," and what you'll check tomorrow. Then you're working with what happened, not what your head is running.
See both setups, and what the therapist can read afterward
Chiron can say what a general chatbot can’t, because it remembers her and it follows your plan. It never offers a diagnosis; everything it says traces back to her own words or your direction.
The hardest thing about the time between sessions is not knowing. Chiron is built so you don’t have to carry that. It watches so you don’t have to, and it tells you the moment something needs you. Underneath every conversation sits a stack of limits a general chatbot simply does not have.
Chiron only ever works inside the modality, voice, and limits you configure.
Every substantive thing it says comes from something you set or your patient said, and the source is visible.
A hard rule enforced in the software itself, not a matter of tone.
A safety layer reads what comes in and what goes out, before your patient ever sees it.
Chiron watches each conversation for signs of suicide risk using the Columbia Protocol (C-SSRS), the standard used in clinical care, and responds by severity.
If a patient is in real danger, Chiron surfaces immediate support and 988, and alerts you at once.
You name what matters for each patient; Chiron tells you the moment it appears.
A standing red team probes Chiron for ways to fail, so the failures are found by us, not by your patient.
Affirmative AI malpractice coverage that names you as an additional insured.†
Rest assured: Chiron is always tending to your patients, and it will reach you the moment you are needed.
The category is loud about what its AI can do. Here is what ours will not.
That judgment is yours, always.
It is a tool in your hands, not a replacement for you.
You can see everything your patient shares with it.
It works toward the goals you set, and nothing outside them.
Your records are yours; cancel anytime and take them with you.
The limits are not the fine print. They are the design.
Chiron is everything a general chatbot isn’t, because it was built to be.
| Chiron | A general chatbot (ChatGPT, etc.) | A consumer companion app | |
|---|---|---|---|
| Configured by your therapist | Yes | No | No |
| Remembers your patient over time | Yes | “Who are you again?” | Maybe? |
| Screens for crisis (C-SSRS) and alerts you | Yes | No | Sometimes |
| Follows your treatment plan | Yes | “What treatment plan?” | No |
| Refuses to diagnose, by design | Yes | No | Sometimes |
| A licensed clinician of record who sees everything | Yes | No | No |
| Malpractice coverage that protects you | Yes | “You must be kidding.” | No |
young people now turn to AI for mental-health support, about as many as see a human professional, and the number keeps climbing. Your patients are already among them. The only question is whose AI they are using.
Cancel any time, and your records leave with you.
No. You remain the clinician of record. Chiron only ever works inside the plan you set, it never diagnoses, and it never delivers care on its own. It extends your work between sessions; it does not stand in for it. And Chiron is yours: you configure and hold the license for each patient, and it cannot be taken or transferred to another therapist. You are the owner.
Yes. Every conversation your patient has with Chiron is visible to you. Nothing is hidden from the treating clinician.
Chiron screens every conversation for risk using the Columbia Protocol (C-SSRS), the standard used in clinical care. If it detects danger, it surfaces immediate support and the 988 line to your patient right away, and alerts you.
Yes. Chiron is HIPAA compliant, and we sign a BAA with your practice.
Chiron carries affirmative AI malpractice coverage that names you as an additional insured. Some malpractice carriers have begun excluding AI-related claims; this is built to cover that gap.
It is yours. Cancel any time and take everything with you. We never sell your patients’ data, and we never use it to train AI.
You don’t have to change how you practice. Configure Chiron for one or two patients, see what it gives back, and grow from there.
Private beta: we’re adding a few practices at a time
admin@167health.com