Real-time loss assessment requires a view of damage during and immediately after an event. This view can be used to assess the building stock, infrastructure and population (‘exposure’) that exists within the flooded area.

Cat model development and validation requires the comparison of modelled event footprints with historical flood footprints, particularly for extent, depth, duration, seasonality, and frequency of occurrence at a location.

Predictive models used can be validated and calibrated via a feedback path in the post-analysis. After an event, initial modelling predictions and associated data are compared to the eventual outcome for loss measurement and event impact. If predictions correlate with the results, the model is verified. If not, the model may be adjusted and re-run against the initial data set for testing and calibration to improve the model for future use.

The data and analytics provided will allow claims professionals to manage and adjust claims remotely, in some instances making early payments to policyholders.

EO satellite data and imagery in the claims process will assist in the initial triaging of loss. They can use them to mobilise third-party service providers and support organisations. In instances of total catastrophic loss, claims can be settled as soon as verified based on EO sourced evidence. REACT provides a shared services platform to multiple organisations, significantly reducing the entry-level cost of real-time flood mapping and claims verification services.

Parametric risk financing schemes require real-time, accurate and trusted data at a given location to trigger such contracts. The platform provides a transparent audit trail/workflow process associated with these contracts to validate data and event triggering. However, human intervention is required to trigger the actual confirmation parametric event.

Users can create alert conditions based on their portfolio or regions. Alerts are triggered once a specified type of event occurs; risk value goes above a certain threshold; the number of properties are affected, etc. The platform then automatically checks these parameters regularly and frequently for the duration of a significant event. When conditions of an alert are reached, the system will notify the user.