Trust

How trust is established, not asserted

Every property record we produce is only as useful as the evidence behind it. This page explains the mechanisms that make that evidence traceable, reviewed, bounded, and defensible — and links to the pages that go deeper on each.

Evidence provenance

Every measurement, image, and derived value in a property record is linked to the capture session that produced it: date, method, equipment, and the person or process responsible. Nothing enters a deliverable without a traceable origin.

Source traceability

Derived outputs — an area, a quantity, a candidate asset — carry a reference back to the raw capture they were computed from. A reviewer or a client can walk backward from any number in a deliverable to the source data it came from.

Review status

Deliverables carry an explicit review state. Model-assisted output is a candidate until a qualified person confirms it. Nothing is released as final without a recorded human reviewer.

Intended-use controls

Every deliverable states the use it was produced for. Using it outside that stated scope — for example, treating a measured-area schedule as a legal survey — is outside the intended use and is called out explicitly on the document itself.

Immutable source records

Raw capture data is not edited in place. Processing produces new derived records that reference the original; the original capture remains available as the record of what was actually observed on the capture date.

Human accountability

Every released deliverable names the person who reviewed and issued it. Automated systems assist; they do not sign off. If something is released without a named reviewer, it should not have gone out and we want to hear about it.

Open standards

Where practical we deliver in open, widely supported formats (point cloud, imagery, spreadsheet, PDF standards) rather than locking evidence into a proprietary viewer, so it can be checked and reused outside our own tools.

No silent AI conclusions

Model output that has not yet been reviewed is labeled as a candidate, never presented as a finding. AI does not approve professional conclusions and does not release deliverables on its own.

Where the boundaries are drawn

Trust also means saying what we don't decide

We document and measure. We do not replace the licensed professionals whose conclusions depend on judgement, liability, and credentialing that sit outside our scope.

Professional boundaries, in brief

Commercial drone operations require appropriately qualified operators under applicable FAA rules.
Legal boundaries and survey reliance require licensed surveying where applicable.
Final tax classifications and return positions require qualified tax review.
Licensed engineering conclusions require qualified engineers.
Thermal anomalies are indicators, not automatic diagnoses.
Coverage and claim decisions belong to insurers and appropriate professionals.
Official valuations require qualified valuation professionals.
Legal opinions belong to counsel.
AI does not approve professional conclusions.

The full statement of these boundaries, by discipline, is on the professional boundaries page.

Data handling and security

How the evidence is held

Provenance and review only matter if the underlying data is handled carefully. Two pages go into detail:

Data handling

What we capture, how it is divided between systems, minimization, retention, and how deletion requests are handled subject to legal and contractual retention.

Read the data handling page →

Security posture

Access control, encryption in transit and at rest, tenant separation, backups, retention controls, incident response, and how to report a security concern.

Read the security page →

Subprocessors and vendor boundaries

Where commercial and technical systems separate

Commercial relationship data (proposals, contracts, invoices, messages, support tickets) is held in SuiteDash, our client relationship and billing platform. Technical property data — captures, spatial data, imagery, and derived records — is held in the Property Intelligence OS workspace and object storage, and is not duplicated into the CRM.

SuiteDash: proposals, contracts, invoicing, messaging, and support tickets
Property Intelligence OS: the property record, review workflow, and client-facing workspace
Object storage: raw and processed capture files (imagery, point clouds, meshes)
Analytics: aggregated, de-identified usage of the public website only

No unreviewed vendor sprawl

We do not add processing vendors that touch property capture data without evaluating their access controls and data handling terms first. A current list of subprocessors that touch client data is available on request during onboarding and under the engagement contract.

Next step

Questions about how a specific engagement will be handled?

Ask before you sign. We would rather explain the boundary now than after a deliverable ships.

Get my property plan Contact us