The whole product is one screen
Most clinical software asks you to learn it. This asks you to press Start. Everything below is the real component, running on invented data, not a picture of one.
Session in progress
In person · 24:10
Simulated session with invented data, not a real patient.
In the room
Live transcript, both voices
On video, the therapist and the patient arrive on separate audio tracks, so each line is attributed with certainty. In person, one microphone hears the room and the note works out who was speaking from context.
Suggestions, not instructions
The copilot reads the last few minutes and offers at most two short prompts. It fires every few segments rather than constantly, because advice that changes every eight seconds is noise.
Crisis language raises a flag
Alerts go to you and only to you. A patient on a join link never sees a risk level, only a supportive message and a crisis line.
Risk language detected · high
Pause and assess directly. If there is imminent risk, follow your local emergency protocol.
Matched: want to die
If there is imminent risk, call your local emergency number.
The note you would have written
Subjective, objective, assessment and plan, plus summary, talking points, observations, impressions and follow-up. Editable everywhere, signed only by you.
A draft until you approve it.
And around it
Off record
One tap stops capture without ending the session. Nothing recorded, nothing transcribed, nothing stored.
Video when you need it
Private rooms with a per-participant token that expires. In-person sessions skip video entirely.
Patient report by email
Approve a note and optionally send a plain-language summary. Clinical impressions stay in your chart.
Questions we get
- Do my patients need an account?
- No. For video sessions you send a link; they type a first name and join. For in-person sessions they do not touch the software at all.
- How does it know who is speaking?
- On video, each participant is on a separate audio track, no voice recognition or enrolment involved, so two similar voices are never confused. In person there is a single microphone, so the live transcript is unattributed and the note works out roles from context.
- What happens if I lose signal mid-session?
- Audio chunks upload independently and retry. A dropped chunk costs a few seconds of transcript, not the session.
- Can I edit the note?
- Yes. Every field is editable before you approve, and the approved version is what is stored.
- Is a BAA included?
- Yes, on every plan. We process protected health information on your behalf, so a business associate agreement is not something we could reasonably charge extra for.
And the half nobody shows you
The patient has an app too
Their own sessions, the ones they booked and the ones they found on the radar, and the note you wrote to them once you signed it. It has no field that could hold your clinical note. That is a property of the query, not of the screen.
What you agreed, on both phones
Homework as the steps you actually agreed, not a scoreboard. Nobody is marked, and nothing is shown to anybody else.
A profile that is dated observations, not prose
Every line traceable to the session it came from. No paragraph of machine writing about a person.
If you need help right now
If you are in immediate danger, call your local emergency number now. This is not an emergency service and nobody here can reach you fast enough. If you can wait a few minutes, the radar has clinicians online this minute and you do not need an account to use it.
No account, no card, no form. You give a first name and you are in a session.