A separate key for every person.
Each user’s coaching data is encrypted with a per-user data-encryption key, derived in a hardware security module from a tenant key-encryption key. Keys live in a UK-region key management service.
01
Continuous youth-MH signal
weekly aggregates · designed against the Code of Practice for Statistics
02
UK sovereign hosting
London region · per-user envelope encryption · UK-resident keys
03
Aggregate only
minimum cell size 50 · k-anonymity · differential privacy ε = 1.0
04
Auditable end-to-end
HMAC-chained access log · SIEM-exportable · documented DPIA
The reality
One in five children in England has a probable mental disorder. Hundreds of thousands are waiting for a first appointment. The most recent national wave was published in November 2023.
Compiled evidence · Three NHS England publications
Youth mental health · England
NHS England · Coverage: England
Last national wave
21 November 2023
Key finding
of 8-to-16-year-olds in England had a probable mental disorder in 2023.1
children waited over a year for CAMHS treatment in 2023/24.2
were still waiting for first contact at end-March 2025.2
note — The most recent national wave was published in November 2023.
1 NHS England, Mental Health of Children and Young People Wave 4 · published 21 November 2023.
2 NHS England · CYPMHS Waiting Times 2023/24 · CYP MH long-read, March 2025.
Measurement cadence
NHS England’s Mental Health of Children and Young People Survey is the gold-standard prevalence study. Since 2020 it has been published as annual follow-up waves to the 2017 baseline. The most recent wave is now two and a half years old, with no successor published.
MHCYP · annual since 2020 (latest 2023)
PULSE · continuous, every week
Pulse is a complement, not a replacement. MHCYP remains the gold-standard prevalence study; Pulse fills the years between waves, and the months after.
How we work together
Each step builds on the last. Every step is independently exit-able. You choose how far to go, and on what evidence.
Pick one policy intervention. We measure whether it changed outcomes in the institutions that received it, against matched controls. Results in months, not years. The easiest place to start.
MHST600+ teams · 52% pupil coverage · target 60% by March 2026 · full coverage by 2029/30
Continuous wellbeing measurement across all 8 early-adopter authorities (Nottingham, Tower Hamlets, County Durham, Manchester, Birmingham, Brighton & Hove, Bristol, Leeds), then onward to the 50 by 2029. Same framework everywhere, real-time outcomes.
Young Futures8 early-adopter authorities opening by March 2026 · 50 hubs by 2029
Aggregated intelligence across every youth-serving institution in your area. Trend data, benchmarking, early warning. The Families First Partnership pathfinders are natural early partners. No individual student data ever leaves the school.
Families First10 LAs in the 2023–25 programme · rolling to all 153 English LAs by March 2027
Poyntr Pulse as a published, accredited national indicator. Open API, transparent methodology, OSR-grade documentation. Cited in parliamentary committees, departmental analysis, and commissioning decisions. Complement to MHCYP, not a replacement.
National indicatorpublished methodology · OSR-grade · FOI-able
Honest by design
We publish.
Methodology, k-anonymity threshold, differential-privacy parameter, suppression rules, every change to any of them. All open, all dated, all version-controlled.
We don’t sell access.
Population statistics drawn from coaching data are a public good. They will not be sold to advertisers, insurers, employers, or hedge funds.
We aggregate.
You see the shape of a region, an age band, an institution type. The minimum cell size is 50 people, well above GSS administrative-data baselines. Smaller, we suppress.
We don’t profile.
No individual student is ever identifiable from a Pulse figure. No re-identification by joining datasets. No targeting of a school, a postcode, or a child by anyone, ever.
We measure.
Whether the policy you funded changed outcomes for the children it was funded for, in the institutions it reached, on a cadence that lets you adjust before the next election cycle.
We don’t replace clinical judgement.
Pulse complements MHCYP and APMS. It is not a diagnosis, not a referral instrument, not a substitute for the clinicians and DSLs who hold statutory responsibility for the child in front of them.
This isn’t a procurement promise.
It’s how the system is built.
What we won’t repeat
Every UK govtech failure has the same lineage: a programme designed for one purpose, an architecture that allowed creep into another, a delayed regulator finding, a quiet cancellation. Each architectural choice in Pulse is built directly against one of these.
Procurement readiness
Every commitment below is in the live codebase today. We do not list certifications we have not earned, frameworks we are not yet listed on, or accreditations we have not yet pursued. Ask us in the briefing where we are with each.
A separate key for every person.
Each user’s coaching data is encrypted with a per-user data-encryption key, derived in a hardware security module from a tenant key-encryption key. Keys live in a UK-region key management service.
Every access, hash-chained.
Each audit log entry is HMAC-chained to the one before it, so any tampering breaks the chain. The chain is exportable to your SIEM in a documented format.
Right to erasure, made real.
Messages, memories, voice, journal entries. We destroy the per-user encryption key, which makes the data permanently unreadable. No undelete. No backup loophole.
Data stays in the UK region.
Application services and the key management service both run in London (europe-west2). KMS HSM-rooted envelope encryption keys are held in-region. No transatlantic data transfer for child data.
Cells of fewer than 50, suppressed.
Population statistics use a minimum cell size of 50, well above the GSS administrative-data baseline of k=5–10 and above ONS small-number suppression thresholds. Differential privacy is applied at ε = 1.0, the conservative end of the published range.
Documented for OSR-grade scrutiny.
Methodology, sources, limitations, and disclosure controls are version-controlled and published. Any change is dated. Designed against the Code of Practice for Statistics from day one.
Get in touch
We will share the full data governance pack, the published methodology, our DPIA template, and a working dashboard tailored to the question you most want answered.