Amazon Comprehend custom entities and custom classifier enable Amazon Kendra search powered by Custom Document Enrichment (CDE) to help end-users perform better searches on structured/unstructured data by filtering and faceting on custom metadata fields.

<div>
<p>This article discusses how to intelligently process insurance claims using Amazon Kendra and custom metadata generated by Amazon Comprehend.</p>
<p>Specifically, the article covers:</p>
<ul>
<li>Training Amazon Comprehend custom classifiers and entity recognizers to identify document types (e.g. home, auto, life insurance) and extract relevant entities (e.g. insurance company, policy number)</li>
<li>Creating Amazon Comprehend endpoints for real-time classification and entity detection</li>
<li>Developing a Lambda function to invoke Comprehend endpoints and enrich documents with custom metadata during Kendra ingestion</li>
<li>Creating a Kendra index, ingesting documents, and using the custom metadata to enable filtering and faceted search</li>
<li>Conclusion highlighting the benefits of using custom Comprehend models with Kendra for intelligent document processing</li>
</ul>
</div>


Use custom metadata created by Amazon Comprehend to intelligently process insurance claims using Amazon Kendra

Related articles