Configuring External File Ingestion

Prerequisites:

Before configuring External File Ingestion:

1. Ensure Siebel Intelligent Search is enabled and configured.
2. Ensure OpenSearch is configured and operational.
3. Enable RAG if the ingested content will be used in RAG experiences.
4. Identify a file location containing only approved files for ingestion.
5. Verify that file formats, sizes, and types are supported by the existing Siebel File Attachments Search ingestion framework.
6. Review content ownership, classification, and retention requirements before indexing documents.

Configure the OpenSearch Engine driver parameters for the external file source.

Parameter	Example or Value	Description
File Ingestion Path	C:\src\LocalFiles	Directory that contains the external files to be imported for indexing.
UploadFile	True/False	Set to True to import files from the configured file location into the Siebel File System staging location.

This Runtime Event is triggered when the UploadFile driver parameter is set to True. It imports files from the local file system and stores them in the Siebel File System.

For example, if the File Ingestion Path is set to C:\src\LocalFiles, all files in this location are imported into the Siebel File System when the Runtime Eventis triggered. The files will be stored in a File Ingestion directory created under C:\fs\opensearchindx\fileindex.

After the files are imported, start indexing for the external file ingestion category in OpenSearch. The indexed content can then be returned by Siebel Intelligent Search and RAG experiences that use OpenSearch as the knowledge base.

Repository and Seed Changes

To enable the ingestion of external files into the OpenSearch engine, a new category is created and mapped to a newly created Virtual Business Component (BC) under the OpenSearch Engine.

Category Name	BC Name
File Ingestion ModernSearch	File Ingestion VBC

Configuring modernsearchconfig.yaml

Configure the File Ingestion category in the modernsearchconfig.yaml file.

Upstream Configuration

upstream:
  - File Ingestion ModernSearch:
      embeddingData: '{{File Name}} is having content {{Description}}'

Downstream Configuration

downstream:
  - fileingestion:
      isIndexed: false
      isAttachment: false

Once indexing is initiated for the File Ingestion ModernSearch category, the system automatically retrieves all files from C:\fs\opensearchindx\fileindex\File Ingestion and indexes their content into OpenSearch.