3BrainAI

For construction finance & real-estate collateral

Monday morning.
A clear review position.

The manager arrives with the report, its supporting records and one clear answer: what has been resolved, what remains open and who owns the next step. Construction Risk Intelligence (CRI) is being built to make that review position traceable.

Request access to The BriefView a public Evidence Pack

The Brief is a private walkthrough of prepared review situations. Editorial evaluation access is free and individually reviewed; other professional access is agreed separately.

From a real place to a clear review position

Historical aerial context and source

Real historical context. Smíchov City, Prague · aerial product reference 2025; exact acquisition day unverified. © ČÚZK / Zeměměřický úřad. Source and image limits

Product context / Current stage: controlled-case Proof of Concept. Institution-specific configuration, calibration and evaluation follow an agreed pilot scope.

Who CRI is for

Construction lenders and real-estate collateral reviewers who need a traceable basis for human review.

What CRI produces

A short, versioned Evidence Pack that keeps the question, sources, uncertainty and next responsible step together.

Where CRI stands

Controlled-case Proof of Concept. Institution-specific work begins with an Evidence Readiness Check.

The continuation of the homepage story

The question is resolved.
The basis stays visible.

The useful outcome is a defensible review position: the resolved scope is clear, and anything outside it remains open.

01 / The unresolved question

A declaration is on the desk.

Drainage and paving are declared ready. The earlier accepted technical record covers earthworks only. The manager needs the adviser’s scoped response.

Pending · response required
02 / The useful report

The answer travels with its basis.

After the model adviser’s response is reviewed and accepted, the report records the resolved scope, its exclusions and the linked references. The manager can explain the position at the meeting.

Scoped acceptance · model
See what supports this model outcome

Real source context: the historical aerial view identifies the setting. It does not confirm drainage or paving.

Fictional records: SMI-M01 defines the scope, SMI-M02 the earlier earthworks acceptance and SMI-M03 the new declaration. SMI-M04 is a scripted adviser reply, not a real inspection.

Analytical proposal: request the dated reply for the same phase and revision; withhold imagery-based completion.

Human acceptance: the model reviewer accepts only the defined external-works update. A bank’s own decision and issued record remain separate.

Illustrative Smíchov review scenario. No actual financing arrangement, site inspection or construction finding is assigned to the real project.

Satellite observations and project records

How satellite evidence supports a bank review

3BrainAI is developing Construction Risk Intelligence (CRI) to bring dated satellite observations together with authorised project records in a versioned Evidence Pack. For a construction lender or collateral reviewer, the useful question is what the available evidence supports, what remains uncertain and which document or expert response is needed next.

Satellite imagery can provide dated context about visible changes at a site. It cannot, by itself, establish engineering completion, contractual acceptance or eligibility for a loan drawdown. CRI is currently in controlled-case Proof of Concept validation; institution-specific use requires an agreed scope and human review.

Explore bank review use cases

Inspect the public Fischamend Evidence Pack

A separate professional section

Go inside The Brief.

Inspect the records, the reasoning and the role of human review.

Free editorial evaluation · Access reviewed individually.

THE BRIEF / PRIVATE PROFESSIONAL SECTION

Explore the reasoning behind the report.

Nine prepared model situations across Austria, Czechia and the Netherlands let you inspect the records, examine the question, reveal a scoped reply and see what human acceptance changes. They are prepared cases, not customer deployments.

Editorial evaluation access to The Brief is free and individually reviewed. Other professional access terms are agreed separately before an invitation.

Real locations and dated source context remain distinct from fictional documents and model review actions.

Request access to The Brief
  1. Professional relevanceWe review your role, organisation and intended evaluation purpose.
  2. Confirm the evaluation scopeWe agree the purpose, permitted use and duration. Commercial terms, where applicable, are handled separately.
  3. Individual invitationYou enter the protected section under the agreed conditions. Submitting a request does not grant access.

No public sign-up or payment is taken on this page. The request is reviewed by a person.

The Brief

A private walkthrough of prepared situations, free for individually reviewed editorial evaluation.

Review the access process

The bank demonstrator is intended for banking professionals; technical advisers participate through the bank’s process. Brief access does not purchase a bank pilot or imply acceptance of your institution’s case.

The four distinct meanings in a record

The review grammar

From context to a responsible review state.

The four meanings remain distinct throughout the record.

  1. 01 · Declared

    Review context

    What the synthetic or institution-provided question says should be tested.

  2. 02 · Observed

    Dated evidence

    What the controlled source record supports within a stated observation boundary.

  3. 03 · Governed

    Uncertainty and reasons

    What remains partial, unverified or outside the permitted inference.

  4. 04 · Output

    Human-review state

    A reviewable record, not an autonomous credit or payment decision.

Two bank questions. One evidence discipline.

Use the record where the question arises.

Construction finance

Examine a milestone declaration against the relevant baseline and technical records. Keep the scope of a previous acceptance distinct from the work now presented for review.

Collateral & portfolio review

Focus follow-up between formal inspections. Ask whether a dated change is relevant to the monitored asset or one of its dependencies, and whether the next step needs documents, an explanation or a site visit.

Imagery supplies dated, surface-visible context. It does not establish property value, legal title, occupancy, hidden condition or protection against theft.

How CRI fits existing controls

Institutional fit

Keep existing controls.
Bring better prepared evidence.

The workflow is designed to add a versioned evidence state between formal control points.

Source dates and reporting cadence depend on the agreed question; the Monday meeting is a narrative example.

The record you can inspect

A concise output with an explicit basis.

Read the logic of a record: how its question, evidence, limits and human review fit together.

Explore the six parts of the record
Evidence PackReading guide · schematic
01 · Start with the question

What is being reviewed — and within which scope?

02 · Declared context

The question, baseline and document authority.

03 · Observed evidence

The dated source and what it actually supports.

04 · Bounded analytical conclusion

Connect the evidence to the question. State only what the available basis supports.

05 · Limits stay visible

Identify missing evidence, uncertainty and what cannot be inferred.

06 · Human review is a separate action

A named reviewer can challenge or accept a claim within an explicit scope. The analytical proposal does not grant acceptance.

A schematic guide to reading a record. No case result or bank decision is represented.

Record fields and status meanings
Evidence Observation, bounded conclusion and uncertainty.
Context Source, observation date and pack version.
Review Reason codes, review state and next question.

Pending means the review is unfinished. Limited evidence means the basis is insufficient. Scoped acceptance records a human’s acceptance within a defined scope. Colour directs attention; it does not make a bank decision.

The target evidence chain

Target architecture

From controlled inputs to a versioned Evidence Pack.

The source, analytical candidate and responsible human review have different roles.

  1. Target workflow Project + observation inputs Controlled context, reports and observation media.
  2. Target workflow Candidate evidence state Bounded findings, reason codes and uncertainty.
  3. Target workflow Human review Confirm, refine or reject the candidate state.
  4. Prototype format Versioned Evidence Pack Controlled review record with visible provenance and review state.

Evidence suitability

A gap remains a gap.

Optical & aerial evidence

Useful surface context depends on resolution, coverage, season and the actual review question. An aerial image alone cannot confirm a milestone.

Radar & documents

Radar may add a different observation capability where it is suitable. Technical and project records remain necessary for questions an image cannot answer.

No data / no change

Insufficient observation cannot establish that nothing changed. The pack keeps the gap and the required follow-up visible.

Historical image: source, dates and limits

Source context

A real place. A model review.

The aerial view is a real historical source, used for orientation.

Smíchov City, Prague · aerial product reference 2025; exact acquisition day unverified. © ČÚZK / Zeměměřický úřad. Reprojected orthophoto prepared for this page. CC BY 4.0.

The real orthophoto was reprojected and resampled for display. It is used for orientation only: the fictional Phase A is not mapped, and the image cannot confirm its milestone.

ČÚZK orthophoto service · Recorded product-year feed

Start with a useful question

Explore the review.
Then define the work.

Use the public record to understand the output, The Brief to inspect the prepared review experience, or an Evidence Readiness Check to discuss your own case.

Explore The Brief

Tell us your role and evaluation purpose. Editorial evaluation access is free and individually reviewed; other professional access is agreed separately.

Request access to The Brief