AWS Launches Machine Learning Search Service Amazon Kendra
In preview since December, Amazon Kendra, an enterprise search tool from Amazon Web Services (AWS) designed to recognize and accurately reply to natural language queries, is now generally available.
Amazon Kendra's aim is to make search more intelligent by recognizing simply phrased queries rather than complex or "just right" combinations of keywords. It also promises to simplify search for enterprises with large volumes of unstructured data, or data that's scattered across multiple silos.
"With just a few clicks, Amazon Kendra enables organizations to index structured and unstructured data stored in different backends, such as file systems, applications, Intranet, and relational databases," AWS said in a blog post this week announcing the product's launch. "As you would expect, all data is encrypted in flight using HTTPS, and can be encrypted at rest with AWS Key Management Service (KMS)."
Amazon Kendra can run via an API or as a console application. It can understand commonly used terms from 14 industries. "Chemical, Energy, Finance, Insurance, IT, and Pharmaceuticals" were already supported in the preview, while "Automotive, Health, HR, Legal, Media and Entertainment, News, Telecom, Travel and Leisure" are new in the general availability release.
Also new with the general availability release are connectors for three additional cloud-based data sources: Salesforce, OneDrive and ServiceNow. That's on top of the SharePoint Online, Amazon S3 and Amazon RDS connectors that were included in the preview.
Amazon Kendra is available out of the North Virginia, Oregon and Ireland regions. It comes in two versions: the production-ready Enterprise Edition and the Developer Edition for test environments. The Developer Edition doesn't support being run on multiple availability zones and limits users to "4,000 queries per day, on up to 10,000 searchable documents across 5 data sources."
The Enterprise Edition, however, can support up to "500,000 documents, and run up to 40,000 queries per day for $7 per hour." Users can also add capacity; an extra 40,000 queries or 500,000 searchable documents will incur $3.50 per hour.
More information on Amazon Kendra is available here.
Gladys Rama is the senior site producer for Redmondmag.com, RCPmag.com and MCPmag.com.