Glance
From the caseload: who you are seeing, when you last met, and whether anything has come in since.
For clinicians
A continuity layer around your work — not another system to keep
Your caseload, and the four questions you have before every session: who am I seeing, what has happened since I last saw them, what should I remember from last time, and is there anything they want to bring into this hour. Healing Space answers those from what the patient chose to share, and deliberately stops there.
Session Prep, in four depths
Read as much as you need and no more — but never be asked to trust a summary you cannot open.
Brief
The interval since the last confirmed session — what the patient shared, the agenda they wrote, what you agreed last time and, where the patient shares them, mood, sleep and questionnaire movement.
Evidence, then the source
Each line is backed by the entry behind it, and the original entry — whole, dated, attributed — is one step further. No item exists only as a summary of itself.
Why the interval is the unit
"Since we last spoke" should mean something specific.
Anchored to a confirmed session
Not an arbitrary last-thirty-entries window. You confirm whether a session went ahead from the same screen, so the interval the record is organised around stays true.
Assembled by retrieval
The continuity surfaces are counts, dates and comparisons over recorded entries, each carrying its date and the person who recorded it — computed, not composed.
And it holds across months
What a patient shared, how often, and how a measure moved between two dates — across the whole course of work. If the record does not contain it, you get nothing back rather than a guess.
What reaches you, and what does not
The boundary is not an inconvenience to work around. A patient who cannot keep something private stops writing anything down.
Authorised context only
You see the briefing the patient approves, from two hours before the session, and your own records: your notes, the summaries you issue and the patient's response to them. Nothing else the patient writes reaches you.
Enforced centrally
Access runs through one authorisation policy across every patient route: your assigned patients and nobody else’s. Sensitive clinician views, such as Session Prep, are logged, and developers hold no standing access to clinical data.
Not the conversation
You do not see what the patient says to Healing Space, what it remembers for them, or what they have said they want help with. Raw conversation is not a clinician feed, and nothing the patient says or does notifies you.
After the session
Preparation is half of continuity. The other half is what you carry out of the room.
Record the session
Confirm it went ahead, keep your note, and — if you choose — write a short summary for the patient and what the two of you agreed.
In your words, attributed to you
The patient sees what you shared with them as you wrote it, marked as yours. Their conversation can refer back to it, and it is never presented as something the software concluded.
The next interval starts there
The next Session Prep begins from that session, so "since we last met" keeps meaning something specific.
Where AI is involved, and what governs it
Specific rather than reassuring — you should be able to tell a patient exactly what happens.
The conversation uses a model
A patient can type to Healing Space between sessions; that text is processed by our AI provider (currently Groq-hosted infrastructure, named in the privacy policy, and due to be compared with alternatives before any pilot). Model training on patient conversations is disabled. The conversation stays on the patient’s side of the platform.
Consequential writes are governed
Where conversation identifies something the patient clearly stated — a mood rating, hours slept, an item for the next session — it is validated deterministically against the patient’s own words and written only after the patient confirms it. A model does not get authority over structured clinical data.
Restrained, and no model writes your brief
The conversation may offer a patient one of five self-help supports, from a list the application controls, and holds back far more often than it offers. Session Prep and the longitudinal surfaces are retrieval and comparison over recorded entries, not model output. No automated decision is made about anyone’s care.
Boundaries, stated plainly
The platform informs your judgement. It never substitutes for it, and it does not quietly take on work you are accountable for.
No monitoring, no alerts
Healing Space is not a monitoring service and raises no alerts: nothing a patient says, answers or does notifies you. If someone talks about being in danger, the AI points them to real-world help such as 999, Samaritans or someone they trust. Safety remains with you and your service, as it does today; the Trust Centre has the detail.
It does not interpret
No causal inference, no judgement about whether therapy is working, and no matching of separate experiences into one recurring issue. Those are clinical judgements.
It is not an EHR
This is a lightweight layer around existing clinical work, not a practice-management system and not somewhere your service is expected to move its whole operation.
Planned, not available today
Where the clinician side is heading. Labelled planned because it is.
Session documentation support
Tools to make documenting a session lighter work, with your professional notes remaining yours and under your control.
Optional session recording
Only as an option, only with explicit agreement from both you and the patient, and only once governance allows it.
Goals and progress, reviewed together
Goals agreed in therapy carried into Session Prep with their history kept attributable, and talking instead of typing for patients once voice is authorised for production.
Where this is honestly up to
We think preserving therapeutic context is worth doing. We have not shown that it makes therapy work better.
No outcome evidence
There is no published evidence for this platform, no independent clinical validation, and no NHS approval. Nothing on this site should be read as claiming any of them.
In early evaluation
We are demonstrating Healing Space — with synthetic data only — to clinicians and to people who evaluate health technology, and seeking evaluation partners. Those are working conversations, not endorsements, and we name nobody for credibility.
Disagreement is useful
If you think the premise is wrong, or that a part of it would get in your way, that is genuinely more useful to us than polite interest.
One platform, seen from four sides
The same intervention, described for the people who use it differently.
See it working — then tell us where it is wrong
Thirty minutes, with the founder, no sales team in the room. Bring your scepticism: professional feedback, a challenge to the approach, or a conversation about evaluating it in your service are all useful outcomes.