01
Answers the list you already publish
Therapies offered, session fees and any concessionary places, insurer cover, online or in the room, where the waiting list stands — straight from pages you approved.
AI chatbot for psychologists and therapy clinics — enquiries, not therapy.
Most first contact with a therapy practice is admin wrapped around a hard decision: which therapies you offer, what a session costs, whether the list is open, whether you take an insurer. Adoomi answers that from the pages you approved, at the hour it is asked and in the enquirer's language. It does not counsel, assess anyone or decide what would suit them. Where a message sounds like crisis it surfaces the guidance your own page publishes and routes the conversation to the person you named.
Never therapeutic adviceYour crisis wording, surfacedEU-hosted, GDPR export and deletion
Fig. 01.A · Psychologists / therapy clinics24/7
Answers when people actually ask
Published
The only source it may answer from
Named person
Takes anything that sounds like crisis
Fig. 02.A · pain quotesThe everyday problem
People take a long time to send the first message, and they send it late
Someone deciding whether to start therapy reads your therapies page, your fees page and your team page in one sitting, usually after everyone has gone home. What they cannot find, they ask — and a therapy enquiry is not a plumbing enquiry. It took something to send, and silence until Tuesday reads as an answer. The practice cannot safely reply from a phone between clients either, because the wrong sentence in that thread costs more than a slow one.
Web07:42
“What therapy modalities do you offer (CBT, EMDR, psychodynamic, integrative)?”
WhatsApp21:15
“How much is a session and is there a sliding-scale option?”
IG DMSun
“How long is the wait for a first appointment, and can sessions be in the evening?”
What you get
What it lifts off intake, and where it stops dead
02
Nobody types a history to get an appointment
What kind of work someone wants, in what format, which days work and a way to get back to them. That is the whole ask: no symptoms, no history, nothing rated.
03
Stops the moment it turns clinical
Fit, suitability and distress end the assistant's turn: your published crisis wording is surfaced and the enquiry reaches the person you named, in the enquirer's own words.
Question coverage
What people ask before they make first contact
The pre-enquiry list — therapies, fees, concessionary places, insurers, format and how long the wait is — answered from your own pages, with fit, history and anything clinical left to your clinicians.
Therapies offeredSession feesConcessionary placesInsurer coverWaiting listOnline or in the roomHow to get started
a
Adoomi
ChatVoice
How much is a session, and do you keep any lower-cost places?
Our fees page sets out the session fee and the concessionary places the practice holds. I quote it as written and never estimate — if your insurer is not named on that page I will ask the team rather than guess.
I've been feeling really low tonight and I don't think I can wait.
I'm not able to help with how you're feeling, and I won't try. Our support page asks anyone who needs help now to ring Samaritans on 116 123 at any hour, or NHS 111 and choose option 2, or 999 if you are in immediate danger. I'm passing this to our named contact now.
Insurer coverOnline or in the roomGetting started
Type a question…● Voice
Buyer intent
What intake needs to know, asked without asking for symptoms
An enquiry is only useful to intake when it says what kind of work someone is after, in what format, which days would work and how to reach them. Adoomi asks exactly those, in your own wording, and stops: nobody is asked to type out their history or how bad things have got in order to get an appointment. Any therapy named back to them is one your pages already list, never a suggestion. What reaches your coordinator is a first call that can actually be made.
Vague enquiries, no context.
Your team chases the basics before they can reply.
a
Adoomi
What they are asking forIndividual therapy, first enquiry
Therapy namedEMDR — listed on your therapies page
Format and timesOnline, or evenings in the room
Contactcaptured@…
3 of 4 captured · 1 in progress
With AdoomiNew enquiry · readyEnq 3013
What they are asking forIndividual therapy, first enquiry
Therapy namedEMDR — listed on your therapies page
Format and timesOnline, or evenings in the room
Contactcaptured@…
One clean handoff.
Your team opens the enquiry and replies — no chasing.
STEP · 01
Ask what they are looking for
Individual, couples or young people, as your site lists it
STEP · 02
Leave the clinical alone
No history taken, nothing rated, nothing reassured
STEP · 03
Take the arrangements
Online or in the room, workable times, insurer or self-funding
STEP · 04
Send it where you said
Distress to your named contact, the rest to intake
Fig. 06.A · after-hoursOut-of-hours
Evenings and Sunday nights, when the message finally gets sent
Adoomi keeps answering the published list after the practice closes: therapies offered, what sessions cost, online or in the room, where the waiting list stands, how someone starts. It does not pretend anyone is awake. An enquiry captured at midnight is an enquiry, not a reply — the alert reaches whoever you set, and a person answers in the morning. What it will not do at that hour is discuss how someone is feeling: that gets your published crisis wording and a route to your named contact.
20:40
“"Do you offer EMDR, and who would I be seeing?"”
Answered from your therapies page
21:15
“"What does a session cost, and is the list open?"”
Fees and list status, exactly as published
22:30
“"I'm struggling to keep going tonight."”
Not triaged: your crisis wording surfaced
08:00
“Intake opens”
Briefed enquiries, and the hard one already with your named contact
Time back
The same paragraph, lifted off whoever is between clients
Every therapy practice answers one list on repeat — therapies offered, fees and any concessionary places, insurers, online or in person, how long the wait is, how to start — and all of it is already published. Adoomi answers from there, so nobody between sessions is retyping it. What still reaches a human is what needed one: fit, an awkward arrangement, an enquirer who is struggling. Each arrives in their own words, by email alert or WhatsApp handoff to whoever you set.
Before Adoomi
- The same questions — which therapies, what a session costs, whether the list is open — get typed out again by whoever is between clients.
- Enquiries land late at night and on Sunday evenings, when nobody is at the desk and the person asking has finally worked up to it.
- A message that mentions how bad things have got waits alongside an invoice query until somebody opens it.
- People who were ready to start ring round several practices and go with whoever replies first.
With Adoomi
- Published answers — therapies, fees, insurer cover, waiting-list status — go out at the hour they are asked, in the enquirer's language.
- Enquiries arrive naming the work wanted, the format, workable times and a way to make contact, with no symptoms typed into a chat window.
- Anything that sounds like distress surfaces your own crisis wording and reaches the person you named, not the shared inbox.
- Intake starts the morning with a briefed list instead of a voicemail box and a scroll back through messages.
Fig. 08.A · controlsTrust & safety
Enquiry questions only — and what clients type is treated as sensitive
The line sits well before anything clinical. Adoomi answers what your pages publish — therapies, fees, insurers, availability — and declines the rest: no advice, no view on risk, no opinion on what suits anyone. A crisis message is never engaged with: it repeats your own published wording — your out-of-hours route, 111 or 999 — and goes to your named contact. Chats are held in the EU, exportable and erasable under GDPR, with logs carrying no content. That is data handling, not compliance.
01
Only what your pages already say
Therapies, fees, insurer cover and waiting-list status come from content you approved. Anything outside it is declined, not improvised.
02
Distress is never worked with in chat
The assistant does not counsel, weigh risk or ask anyone to explain more. It repeats the crisis wording you publish and hands the conversation to your named contact.
03
Fit stays with a clinician
Which therapy suits someone, whether a therapist is the right match, anything about their history — each one is a handover. The assistant captures the enquiry and stops.
04
What clients type stays sensitive
Chats sit on EU servers, come back out or get wiped on request, and their content never reaches logs — only IDs do. Compliance stays your practice's job.
How it works
How an AI chatbot for psychologists and therapy clinics works.
01
Connect content
Point it at your therapies, fees and getting-started pages. Nothing outside that approved content can become an answer, so a tightly written site makes a safer assistant.
02
Add the snippet
Set the destinations: ordinary enquiries to intake, anything that sounds like crisis to a named person, with your own published crisis wording surfaced first.
03
Test & go live
Read what it proposes to say, put your hardest message to it yourself, then add one line to your website. The wording gets tightened once real enquiries start arriving.
FAQ
FAQs about AI chatbots for psychologists and therapy clinics.
The assistant answers both from your own pages — the therapies you list, what an assessment session covers, online or in the room. It repeats your wording rather than describing therapy in general, and it never suggests which one an enquirer should have.
Crisis messages are never worked with in chat. The assistant stops, repeats the crisis guidance your own page publishes, and routes the conversation to the person you named rather than the general enquiry inbox. It does not assess risk, because assessing risk is clinical work.
Matching is a clinician's judgement and stays there. The assistant repeats the specialisms your team pages publish, captures what the enquirer is looking for, and hands the fit question over. It will not confirm a therapist has space unless your pages say so.
Client messages are treated as sensitive from the first line. Conversations stay on EU servers, and a client can ask for theirs to be handed over or wiped under GDPR. The words themselves never reach a log — an enquiry is recorded by identifier only. No tracking cookies are set.
Waiting-list status is answered exactly as your page states it and no further. The assistant holds no diary and cannot book, hold or move a session — it captures the request and passes it to intake, who arrange it the way they always have.
Pricing starts at £0 a month, which covers 50 message credits while a practice tries it out. Above that, £8 buys 200 credits, £20 buys 500, £75 buys 4,000. Set-up is ~10 min of work — connect the pages, read what it drafts, add one line — and nothing is live until the practice signs it off.
Why this matters
Built for the enquiry, silent about the room
The risk for a therapy practice is not an assistant that is unhelpful — it is one that is helpful in the wrong place: reassuring someone, ranking therapies, judging how bad things sound. Adoomi holds no view. It answers from your approved pages, in the enquirer's language, whenever asked, and treats fit, distress and history as handovers rather than as questions to have a go at. Those are limits in how it was built, not promises: silent page, no answer; no fee guessed; nobody's state judged.
Cover for the hours people actually enquire
Therapy questions get asked late. The published answer — therapies, fees, format, list status — goes out then, in the enquirer's language, rather than on Monday.
The only source an answer may come from
Fees, concessionary places and insurer cover are repeated as your pages word them. Where the pages are silent, the assistant says so and captures the question.
Where a distressed message goes
A message that sounds like crisis stops the assistant, surfaces your own published wording and reaches the individual you named instead of the shared inbox.
Start free
Try it on the message you would least like answered badly
The therapies, fees and getting-started pages are kept accurate already — that is the whole of what it may draw on. Say who picks up an ordinary enquiry and who is named for anything that sounds like crisis, then send it the message you would dread finding on a Monday and read what it says back. Setting it up is ~10 min of work, £0 and 50 message credits are enough to judge it by, and no client sees a word until you have agreed to every one.
No card neededYou approve every answerCancel anytime
Fig. 12.A · Psychologists / therapy clinicsExplore
More Healthcare chatbots
- AI chatbot for Cosmetic clinicsSee how Adoomi handles enquiries for cosmetic clinics.
- AI chatbot for Dental practicesSee how Adoomi handles enquiries for dental practices.
- AI chatbot for PharmaciesSee how Adoomi handles enquiries for pharmacies.