Insurance companies

Risk assessment based on actual property condition.

Room-level condition scoring and image analysis give you an objective, standardised basis for premium setting and claims risk assessment, not just what the register says.

Condition scoring

Every room scored. Consistently. At scale.

Vision Engine™ analyses uploaded property images and returns a standardised condition score for each area of the property: kitchen, bathroom, interior and exterior assessed separately.

  • Kitchen

    C2

    Good condition

    Modern appliances, updated surfaces. No visible wear or damage. Low claims risk indicator.

  • Bathroom

    C4

    Fair condition

    Dated fixtures, visible tile wear. Elevated moisture risk. Flag for review at renewal.

  • Interior

    C3

    Average condition

    Standard finish, minor cosmetic wear. Consistent with age of building. Normal risk profile.

  • Exterior

    C1

    Excellent condition

    Recently maintained. No visible damage or weathering. Lowest risk category.

Where condition scoring changes your decisions.

  • Underwriting

    Risk-based premium setting

    Incorporate room-level condition scores into your premium models as structured input variables. Properties with elevated bathroom or moisture risk scores can be flagged, tiered or declined at point of application, automatically.

  • Portfolio review

    Renewal risk screening

    At renewal, request updated images and re-score the property. Identify properties where condition has deteriorated since underwriting and adjust premiums accordingly, before a claim occurs.

  • Claims management

    Pre-claim condition baseline

    Establish a documented condition baseline at policy inception. In the event of a claim, condition scores provide objective evidence of the property's state before the incident reducing disputes.

  • Data quality

    Beyond the register

    Official records rarely reflect actual property condition. A property built in 1960 may have been fully renovated last year or may be in serious disrepair. Condition scoring gives you the real picture.

Integration

Fits into your existing premium models.

Condition scoring is delivered as structured JSON via API and can be mapped directly to your existing variables. We provide documentation and mapping examples at onboarding.

  • REST API

    Structured JSON output

    Condition scores per room, overall property score, confidence indicator and raw feature flags. All returned as structured JSON, ready to feed directly into your premium calculation engine.

  • Image formats

    JPEG, PNG and HEIC

    Images can be submitted via API or uploaded in Nezto Pro. No specialist camera equipment required, standard smartphone photography produces reliable condition scores.

  • Data sources

    Images + registry data

    Condition scoring works from images alone. When combined with market data, the output is enriched with building age, area characteristics and local risk context.

Common questions

  • How is image analysis used for claims risk assessment?

    Vision Engine™ analyses uploaded images and returns condition scores per room: kitchen, bathroom, interior and exterior are assessed separately on a standardised scale. The results can be used as input to your own risk models to identify properties with elevated claims risk prior to underwriting or at renewal.

  • Can condition scoring be integrated into existing premium models?

    Yes. Condition scoring is delivered as structured JSON via API and can be mapped directly to your existing variables. We provide documentation and mapping examples for the most common premium model architectures at onboarding.

  • What image formats and data sources are supported?

    Nezto supports JPEG, PNG and HEIC for image analysis. Images can be submitted via API or uploaded directly in Nezto Pro. For market data, the platform aggregates sources such as SCB, Boverket and Lantmäteriet, as well as proprietary transaction data from broker partners.

  • How reliable is the scoring without professional photography?

    Standard smartphone photography is sufficient for reliable condition scoring. The model is trained on a wide range of image quality levels. A confidence score is returned with each result, indicating where image quality may affect reliability.

Underwrite on what you can actually see.

Book a demo and see condition scoring run on real properties from your portfolio.