Analytics

Here are the Niseko release highlights for the Guidewire Analytics products. For more details, see the documentation for Analytics Manager and HazardHub.

Analytics Manager

A validation endpoint is now supported when the workflow system is PolicyCenter

New support for a validation endpoint overcomes a restriction in the PolicyCenter API. Specifically, Analytics Manager transforms PolicyCenter data, at configuration time and at run time, so that you're able to:
  1. Use the validation endpoint to learn the structure of the transformed data.
  2. Apply your knowledge of that data structure to code the request message that will be submitted to an analytic endpoint.

Include more values in the assessment summary

When you've defined an endpoint whose purpose is Create Assessment Summary, you can assign values to a variety of summary attributes whose meaning is specific to your use case. In some cases, you can display the assigned values to the end user. In all cases, the assigned values are stored in the assessment summary on the workflow system, where the values can be inputs to custom processing.

Support for special characters in requests and responses

To enable messages that include Japanese, French, or other languages that use special characters, Analytics Manager now provides the ability to set the message encoding and Content-Type header value to the following value: application/json;charset=UTF-8.

Improved performance through identification of duplicate requests

When two requests are received for the same assessed entity, such as a claim, the processing can happen faster Analytics Manager has been configured to interpret a new request as a duplicate of the last message that was received. That configuration prevents the system from sending the duplicate message to the analytics endpoint.

This capability is especially useful for handling message traffic during times of peak processing. However, the only comparison is between the current message and the previously received message, regardless of the interval between which the two were received.

To improve performance, Analytics Manager can identify a request as a duplicate of the last message it received and ignore that request rather than sending the duplicate to the analytics endpoint. You can specify variables that Analytics Manager can ignore (like timestamp) when comparing messages. (Analytics Manager only compares a message to the previously received message.)

Ability to customize influence and related details on an assessment screen

By default, an assessment screen shows a rightward-expanding green bar as the way to express a Predict feature that adds to an output score. It shows a leftward-expanding red bar as the way to express a feature that subtracts from the output score. Now, you can customize the user interface as follows:
  • Specify the expansion direction that expresses a positive or negative influence. For example, you can specify that a rightward-expanding red bar indicates a negative influence.
  • Substitute a different set of colors.
  • Display or hide the magnitude of a given influence.
  • Display or hide a tooltip that shows the influence detail for a given bar.

You also can display or hide the solution estimate and the range of estimates in the header.

Risk Distribution chart for HazardHub transactions

The Risk Distribution Chart provides insight into the HazardHub transactions that Analytics Manager has processed.

Easy configuration of response settings

You can adjust response settings easily for score bands, as well as for risk categories such as High or Low. You can customize the displayed color, providing quick insights about the business meaning of the response.

HazardHub Wildfire Risk Score 3.0

Wildfire Risk Score 3.0 is a standalone product and comprehensive wildfire risk model. It provides wildfire risk and exposures for every property. Wildfire Risk Score 3.0 provides 3 scores for any address in the United States: a location hazard score, a building exposure score, and a comprehensive wildfire risk score. It also provides all the underlying data behind the scores. This tool lets you underwrite wildfire risk faster and with greater precision and more accurate pricing.

For a short demonstration of this feature, watch this video: