Property evidence usually begins the moment something goes wrong. HomeCheck begins years earlier — an item-level condition record, sealed at the point of capture, that travels with the address across tenancies and arrives as structured data rather than a folder of photographs.
At first notification you can usually establish what is broken. What rarely exists is a dated, trustworthy account of what that item looked like beforehand — which is precisely the fact that decides reserve, betterment and recovery.
The same record that settles deposit disputes between landlords and tenants turns out to answer the questions a claims or underwriting team asks first.
Every item and surface carries a dated condition from before the loss, so a reserve can be set against the actual pre-event state rather than an assumption. Where an item was already worn, that is on the record and was on the record long before anyone had a reason to claim it.
Inspections line up item by item across years and tenancies, so what changed is visible next to what was already there. Gradual wear and a discrete event look different in a continuous record in a way they never do in a single photograph.
Footage is cryptographically signed at the point of recording. It cannot be edited, filtered, or imported from a camera roll — capture has to happen live in the app. Altering a completed record breaks its signature, which is the property that makes it worth relying on.
The record is the same in each case. What changes is the question being asked of it.
Pull the pre-loss condition of the affected rooms and items at notification, so a claim can be routed and reserved against a known starting state instead of waiting on an inspection to establish one.
Aerial and external data describes the building. A HomeCheck history describes the inside of it over time — fabric, fittings and appliances, and how they have actually been maintained across successive tenancies.
Item-level age and condition support a defensible position on betterment, and a dated record of pre-existing damage gives a clear basis for discussion where a claim and the history disagree.
A data pack is a structured record of a property’s inspection history, not a bundle of PDFs. Every row is an item at a point in time, which is what makes it comparable across inspections and ingestible by your systems.
| Field | What it holds | Example |
|---|---|---|
| Property reference | Stable identifier for the address, consistent across every tenancy | HC-32ARD-D6 |
| Inspection date | When the capture took place, recorded at the device at the time | 2025-03-14T16:42:31Z |
| Inspection type | Check-in, check-out, or mid-tenancy | check_out |
| Room | Room the item was recorded in | Bathroom |
| Item | The identified asset, tracked across inspections even when described differently | Walk-In Shower |
| Condition | Condition assigned from the footage at that inspection | Damaged |
| Change vs previous | Whether the item changed since the last inspection, and in which direction | new_damage |
| Pre-existing flag | Whether the condition was already present at the preceding baseline | false |
| Depreciated value | Item-level value adjusted for age and recorded condition | 180.00 GBP |
| Note | The description attached to the item at that inspection | Cracked tile — new vs T2 baseline |
| Media reference | Pointer to the sealed footage the condition was derived from | media/8f2c… |
| Record signature | Signature of the sealed record, for verifying it has not been altered | sha256:4b9e… |
Illustrative field list. The exact schema is agreed during onboarding and versioned, so an integration does not break when fields are added.
Both are available on the Agency plan, and both carry the same record. Which one suits you depends on whether you want a periodic view of a portfolio or an answer at the moment a claim lands.
A machine-readable export of the inspection history for an agreed set of properties, delivered on a schedule or on request.
Request the condition history for a single address at the moment you need it, so pre-loss state can sit inside your own triage workflow.
Answered here rather than in a follow-up call, including the one where the honest answer is not yet.
| Question | Answer |
|---|---|
| Where is data hosted | Microsoft Azure, in European data centres, under a Data Processing Agreement incorporating Standard Contractual Clauses. |
| Encryption | TLS 1.2 or above in transit; AES-256 at rest via Azure Storage Service Encryption. Backups inherit the same protection. |
| Transfers | EU Standard Contractual Clauses, and the UK International Data Transfer Agreement for UK-specific transfers. |
| Controller | HomeCheck Ltd is the data controller for platform data, operating under UK GDPR and EU GDPR. |
| Access control | Role-based access per property; production access limited on a least-privilege basis and logged. Analysis is automated — no one reviews footage as a matter of course. |
| Breach notification | Notifiable breaches reported to the ICO or the Data Protection Commission within 72 hours of becoming aware. |
| Retention | Inspection records for the tenancy plus six years, matching the realistic life of a dispute. |
| Certification | HomeCheck does not currently hold ISO 27001 or SOC 2. We build on certified infrastructure and will state it here, with date and scope, if and when that changes. |
Full detail on our Security page and in the Privacy Notice.
A record is only useful to an insurer if everyone agrees what it does and does not establish. These limits are stated in our Terms and we would rather you read them now than discover them mid-claim.
No structural assessment and no opinion on the value of the building or its fabric.
It does not replace a gas safety record, EICR, EPC or any statutory inspection.
The record shows condition and change. Who is responsible, and under which policy terms, remains your determination.
We separate pre-existing condition from later change. Any inference drawn from that is yours, made under your own processes.
HomeCheck supplies evidence. It does not inspect on your behalf, quantify a loss, or settle anything.
Condition scores come from computer vision. Lighting, obstruction and coverage all affect them, and they should be reviewed before being relied on.
Tell us what you underwrite or handle and we will walk you through a sample data pack, the API, and what an integration would actually involve.
