Amazon Kendra launches confidence scores

[Amazon Kendra](/kendra/) is a highly accurate and easy to use enterprise search service powered by machine learning. Starting today, AWS customers can use confidence scores in Amazon Kendra to improve the search experience and show the most relevant results for user queries. Confidence scores allow customers to see the estimated accuracy of a search result. Amazon Kendra assigns confidence scores into four main categories: Very High, High, Medium, and Low for various document results (e.g. Factoid Answer, FAQ). With confidence scores, developers can better distinguish between returned results and set thresholds for when results should be displayed. By combining confidence scores with Amazon Kendra’s ability to intelligently search unstructured data using natural language, you can return more accurate answers and create a search experience that more closely resembles asking a human expert. Confidence scores for Amazon Kendra is available in all AWS regions where Amazon Kendra is available. To learn more, visit the Amazon Kendra [documentation page](https://docs.aws.amazon.com/kendra/latest/dg/API%5FQueryResultItem.html#Kendra-Type-QueryResultItem-ScoreAttributes).