Why the money cannot reach the verdict.
Independence at HeatAssure™ is built rather than claimed: six constraints on how the record works, each published and each written to be checked. No one who pays can change what a reading says, which is why it still stands up with the people it sometimes disappoints.
Six constraints, held and published.
Each one is a rule the record is built to obey, written so you can check it. Together they mean that nothing here ranks you against anyone else, and no one can pay to change what a reading says.
- The method is fixed and governedThe national standard in force supplies the method, and the home's own design supplies the benchmark. The thresholds that set a verdict change only through a governed, versioned process, never per customer. The calculations are validated against the government's own reference, and the specification is open to any customer, prospective customer or their technical assessor under diligence.
- Every home against its own designA verdict compares a home with what that home was promised, and names the cause the evidence supports: the design, the fabric, the settings, the tariff, the install or the product itself. We publish no ranking of one maker or installer against another, and no one is shown a comparison of work they do not answer for.
- No reading is bought, delayed or buriedA customer funds the record and never edits, delays or suppresses a reading. The homeowner sees their reading no later than anyone who pays, including while it is disputed, and a failing check stays visible until it is fixed.
- The verdict is walled off from the saleA verdict’s thresholds are set by the governed method, never in a commercial conversation, and no commercial term references a verdict’s outcome. Our income is concentrated in a few customers today; the figure is disclosed in diligence, and we will publish it as the customer base broadens.
- Every reading carries a right of replyAnyone whose work a reading touches can put evidence on the record, inside an agreed window, before the verdict is marked settled. The reading is never withheld while it is argued, and a corrected reading is republished with the change visible.
- Readings go to the home’s own peopleThe homeowner sees their home, a manufacturer its own base, a landlord its own stock, each fixed by contract at deployment. No home-level or brand-level register is ever published, and access beyond the home’s own people is granted only by whoever the record is about, for a stated purpose.
It is also why no customer is named on this site: a record-keeper belongs to none of the parties it judges. References and deployment detail are shared in confidence as part of diligence.
Data & IP boundaries.
Customer data stays with the customer
Raw data is processed in a separate environment for each customer, and anything that could identify you, your homes or your products stays inside it. Under each customer’s contract licence the record improves from patterns that carry no maker, product or portfolio attached to them, and that improvement is part of what every customer buys.
Evidence is portable
Evidence packs and audit trails export in human-readable and machine-readable formats.
Leave with everything
Cancelling ends the subscribed service, not your access to the evidence already delivered. The record’s history stays intact.
IP boundaries are explicit
Your raw data stays yours; we hold it as custodian, and export is available at any time. The evidence packs and per-home results we produce for you are licensed to you without restriction, to keep and use in your own systems. The governed record, the verdicts and the method behind them remain Aeterno property.
Security & compliance.
Two things are set out here: what the platform runs on today, and what is configured with you at deployment. Every line can be walked through live.
How the platform runs today.
Separated by customer
Each customer’s data is held apart, with no mixing between customers
Every action is logged
Anything the platform does outward needs an approval first, and lands in a log that cannot be edited afterwards
Audit trails
Action and access histories are recorded for review and export
Data encryption
Encryption at rest and in transit using industry-standard protocols
Configured with you at deployment.
Single sign-on
SSO against your identity provider over SAML 2.0 or OIDC, configured at deployment
Role-based access
Each team sees only the part of the record it needs, mapped to your teams at deployment
Regional hosting
Hosting region and data-localisation requirements agreed per deployment
Where the evidence supports compliance.
MCS
Evidence outputs can support commissioning and installation-quality reviews against the requirements of MCS, the UK's installer certification scheme
Building regulations
Evidence packs can support building regulations compliance workflows
GDPR
Data processing agreements, privacy by design and data subject rights
Due diligence
Documentation supporting customer security review is available on request
What every recommendation shows.
Before any team acts, they can see what triggered the recommendation, what supports it, how confident the platform is, and what happens next. Every input, and the confidence attached to it, is named on the record.
What set it off
What triggered this recommendation? Survey evidence, sensor data, weather, a change in the home: every input is named.
Evidence
What data supports it? Raw readings, derived results and the wider context of the home are linked so nothing is taken on faith.
Confidence
How sure are we? A confidence score with stated bounds, so you see when the platform is uncertain.
Action
What should happen next? A concrete, reversible step, with the undo agreed before anything runs.
- A live governed platform: every screen resolves against a governed viewpoint, and every claim carries its validation status, including the red ones.
The red ones are the claims that do not yet pass. Failure stays visible until it is fixed.
You set the policy.
HeatAssure runs at the level of autonomy you choose. Three settings, Assist, Govern and Auto, decide how much happens before a person is involved.
Every recommendation waits for a person: the platform shows its evidence, your team decides and acts, and every deployment starts here.
Actions run inside bounds agreed in advance. Escalations and edge cases stay with people.
Approved kinds of action run without waiting, still fully logged, and anything that drifts past the limits you agreed is undone automatically.
Changing the setting is a decision you make, against quality measures agreed together and evidence that each kind of action can be undone.
Before procurement asks.
Bring your hardest questions.
We can walk through governance, ownership and deployment in as much detail as your team needs.