Step 5: Configure and publish the processes

1Open the Amazon Bedrock with File process.
2In the Set Flow Configuration step, in the Assignments field, enter values in the following fields:
• - Bucket_Name - The bucket name that you entered in the AmazonS3forAmazonBedrockConnection connection.
• - Wait_Time_to_Create_Data_Source - The idle time to create datasource for knowledge base. If the server has significant delays, you can increase the time.
• - Wait_Time_Ingestion_Job - The idle time for ingestion job. This time depends on the number and size of files to be converted into a vector database. If the files are large and numerous, you can increase the time to reduce the number of status check requests.
• - Embedding_Model - The embedding model in the following format:
arn:aws:bedrock:us-east-1::foundation-model/{Embedding_MODEL_ID}
• - RoleArn - The role arn in the following format:
arn:aws:iam::{USER_ID}:role/{ROLE_NAME}
• - Storage_Configuration - Update the storage configuration prompt instructions using the Expression Editor, as shown in the following sample code:
<storageConfiguration>
<pineconeConfiguration>
<connectionString>{Index Host}</connectionString>
<credentialsSecretArn>{Secret ARN}</credentialsSecretArn>
<fieldMapping>
<metadataField>metadata</metadataField>
<textField>text</textField>
</fieldMapping>
<namespace></namespace>
</pineconeConfiguration>
<type>PINECONE</type>
</storageConfiguration>
If you want to specify a different vector database, specify the appropriate parameters. For more information, see the API_agent_CreateKnowledgeBase documentation.
• - LLM_Model - The LLM model ID.
3 Save and publish the process.
4To publish the Send Email with Process Result process, click Actions in the row that contains the process and select Publish.
5To publish the Write File to Amazon S3 process, click Actions in the row that contains the process and select Publish.