You are here:
Set Up Amazon Kendra for External Storage Search
Configure an Amazon Kendra index to ingest and search files in the Amazon S3 bucket connected to your Salesforce org.
Before you begin:
- Verify that you have permission to create and manage Amazon Kendra indexes, data sources, IAM roles, and related resources.
- Prepare metadata sidecars for files that existed before External Storage Search was enabled. See Prepare Existing Files for Amazon Bedrock.
-
Create an Amazon Kendra index.
See Creating an index - Amazon Kendra in the AWS documentation.
-
Create searchable text for audio, video, and image files that you want Kendra to
index.
Use an appropriate service, such as Amazon Bedrock Data Automation, to create transcripts or summaries. Amazon Kendra doesn't generate transcripts or image summaries. See Amazon Bedrock Data Automation in the AWS documentation.
-
Add the connected Amazon S3 bucket as a data source.
See Amazon S3 - Amazon Kendra in the AWS documentation.
-
For the metadata files prefix, select
__sfdcroot__/__metadata__/. -
Create the StringList custom attribute required to store Salesforce ContentDocument
IDs.
See Custom fields or attributes in the AWS documentation.
- Synchronize the data source.
- Verify that the expected files and metadata are indexed.
Konnten Sie Ihr Problem mithilfe dieses Artikels lösen?
Geben Sie uns Feedback, damit wir uns verbessern können.
