Trauma-informed by design
Meet hard information on your terms.
SafeTeaNet is the trauma-informed layer between a person, their support people, and the systems around them. Hard information arrives on their terms — never censored. The people they trust can stand beside them for the difficult parts. And their story travels with them, shared only when they choose, revocable in one tap.
Not a news app. Not a crisis service. A continuity layer with the person at the centre.
The problem
The world doesn't ask how you'd like to hear it.
A headline arrives engineered to alarm. A search for a diagnosis returns the worst number first. For someone carrying a difficult story, information doesn't just inform — it ambushes. And then every new service compounds it: a new psychologist, a hospital admission, a school meeting, each asking that person to reconstruct their history from the beginning. Neither of those costs is administration. Both take something real, every single time.
“The platform carries the story, so the person does not have to keep retelling it.”
How it works
Built so agency always wins.
Every capability follows the same rules: nothing is hidden, nothing is forced, nothing is shared without a deliberate choice — and the original is always one tap away.
Information without the ambush
News and searches arrive considered against your preferences — graphic detail set aside, statistics given context, persuasive framing named for what it is. Never censored: the original stays one tap away, always.
Human support, before automation
Choose a trusted person who can look at something first or know when you're reading something hard — only ever at your explicit request, seeing only what you send.
Your people, and what they can see
The network the product is named for: the people and services around you on one screen, each showing exactly what they can currently see. Deny-by-default, changed instantly, revoked silently.
Record it once, on your terms
Your story, communication preferences, sensitivities and safety strategies — as much or as little as you choose. Naming a difficult topic never requires describing it.
Brief a new provider without starting again
A purpose-specific briefing built from what you've already recorded, reviewed by you before anything is shared — and explicit about what is deliberately left out.
The idea
Many systems. One person. One continuity layer.
- The individual sits at the centre — their story, preferences, permissions and support network move with them.
- Systems integrate around the person. SafeTeaNet is never the system of record — it is the consent, context and communication layer between them.
- Sharing is deny-by-default: no service and no person sees anything until it is deliberately, specifically granted — and every grant can be revoked in one tap, silently.
- Today the demo proves this with synthetic services — consent-gated imports, every record labelled with its source. Real connectors to health, government and community systems are the roadmap, stated plainly as such.
Privacy by architecture
Trust is structural, not a setting.
Local first
Your profile lives on your device. It is never uploaded, and no AI service ever receives your story — by construction, not by policy.
Deny by default
No person or service can see anything until you grant it — specifically, purposefully, and revocably.
Provenance always
Every record shows where it came from — entered by you, imported from a service, or suggested by AI — and an AI suggestion never becomes fact until you confirm it.
Never monetised
No advertising, no data sale, no behavioural profiling. A platform for trauma-adjacent stories that funded itself on those stories would deserve no trust at all.
Try it
The live demo, in four moves.
Six synthetic personas — a veteran, a neurodivergent parent, a carer, a survivor, a buddy and a provider. Everything runs in your own browser; you can't break anything.
Sign in as Alex and open SafeTea News. Real headlines sit alongside demo stories, each considered against his preferences. Open the veteran article, then “How was this adapted?” — then “Read the original”: nothing is locked away, and his own strategies and his partner Heidi are offered before he decides.
Open My SafeTeaNet and tap Dr Jane Example: she can see exactly two things. Revoke one — gone instantly, silently — then grant it back. Safety on the way in, consent on the way out.
The “New psychologist” briefing shows what a new provider actually receives — including the line “Your trauma history is not included.”
Not a veteran, no buddy by choice. Connect a community service in the Integration Hub and watch the record arrive with its source labelled. Services connect to the person — never the reverse.
Every person, story and record in the demo is synthetic. The demo stores its data in your browser only, and it is not a crisis service — if you are in immediate danger, contact emergency services.
Where it's at
An MVP, stated honestly.
SafeTeaNet is a working MVP: the consent engine, adaptation pipeline, provenance model and integration architecture shown in the demo are real and tested. It has been through a formal adversarial review — privacy, security, trauma-informed design and clinical scepticism — and the findings are fixed or published in the open backlog.
Before any real person's story enters the platform, two gates remain deliberately closed: device-level authentication and encryption at rest. Until then, everything you see runs on synthetic lives.