Actions

Build your view of risk

Risk submissions arrive incomplete containing unstructured and heterogeneous risk data. Digitizing and turning submissions decision-ready requires different actions to be flexibly combined. The Cytora platform enables you to choose from a range of digitization actions to fulfil each schema field providing flexibility to build your own view of risk. Digitization actions include extraction, inference, enrichment, entity resolved and evaluated field types. The combination of multiple digitization actions within a single schema enables larger decision-ready schemas that contain all of the risk data required for decisioning, eliminating manual work, creating capacity and unlocking end to end automation.

File classification

Classify your intake

Classify risks into granular categories that reinforce your view of risk
Multi level category hierarchy enables document types to be classified at a granular level
Define mappings between submission documents received and schemas to provide control and consistency over the source of risk fields

Extraction

Extract key data fields

Natural language configuration with no training required enables you to fulfil your target data structure
Supports all field types and nested schemas for example - extracting a list of policy periods, each containing a list of claims
Multi-modal supporting all input formats, from text based, to tabular, to audio/video

Enrichment

Augment with internal and external data sources

Integration of external data sources to augment risk submission data and automate manual data look ups
Integrate internal data sources to enhance risk context and automate risk workflows
Preintegrated data ecosystem across commercial and speciality insurance

FAQs

Learn how to create decision-ready risks by combining digitization actions

What is a digitization action?

A digitization action is a type of digitization that contributes to decision-ready risks, claims or other transactions. Cytora makes available all digitization actions required in commercial insurance including extraction, enrichment, evaluation, entity resolution, input fields and internal data integration.

Which actions are available in the platform?

The Cytora Platform currently supports classification, extraction, enrichment, entity resolution, input fields, inference and evaluation. Different actions are configured to fulfil different schema fields.

How are classifiers configured?

Classifiers are configured in natural language as the first step following the creation of a digitizer, and the step preceding field definition. In the ‘file types’ tab of your digitizer, name your classifier in the ‘name field’, then explain its function just as you would explain it to a human in the ‘description’ field.

How is field extraction configured?

Configured in exactly the same way as classifiers, using natural language, you can configure field extraction in the ‘fields’ tab of your digitizer. Name your field, describe its function to the platform just as you would explain it to a human, select your output format from a range of options including number, string or dropdown, then test your fields for extraction in the ‘Teach’ interface.

What our customers say

“

We pick better risks and with the speed, save operating expenses. Cytora means we can come to better underwriting decisions, save operating expenses, so we increase our margins."