Adding a data source using connectors

ZSearch uses ZBrain Builder to securely connect to and ingest enterprise data. Administrators can add data sources using built-in connectors to systems such as Google Drive, SharePoint, and other supported repositories. Each connected data source becomes a managed input to ZSearch, forming the basis for searchable enterprise knowledge.

Adding a data source defines where data is sourced from, how it is indexed, and how access permissions are enforced across search and collaboration workflows.

Data source configuration

Data sources are configured from the Data Sources section in ZBrain Builder. Once logged in to ZBrain -

• Click the gear icon in the upper-right corner and select Data Sources.

• This will redirect you to the Data Sources page.

Adding a data source

To add a data source -

• Click Add Data Sources. This shows you the list of data sources available to connect to.

Adding a connector

When adding a new connector, administrators need to specify the connection details required to securely access the source system.

During configuration, administrators:

• Select the connector type (for example, Google Drive or SharePoint)

• Provide the required authentication and connection parameters

• Choose the vector store used for indexing and retrieval. By default, OpenSearch is selected as the vector store.

ZBrain Builder supports multiple vector store options, including Economical, Pinecone, Chroma, and OpenSearch.

Organizations may also configure custom vector store connections, depending on their deployment requirements.

Adding a custom vector store connection

ZBrain Builder allows administrators to configure a custom vector store to manage how embeddings and indexed data are stored for enterprise search and retrieval.

To add a new vector store connection -

• Click 'Add New Connection'

• Once clicked, it opens the Add New Connection dialog

The Add New Connection dialog is used to connect an external or hosted vector database to ZSearch.

This configuration determines where vectorized representations of enterprise content are stored and how they are accessed during indexing and search.

The Add New Connection dialog contains the following fields:

Vector Storage

Select the vector storage provider to use for this connection. This dropdown allows administrators to choose from supported vector store options, such as Pinecone Hosted or Qdrant available configurations.

The selected vector store determines the underlying storage and retrieval behavior for indexed data.

The administrator needs to fill in the following details to establish a connection.

Name

Provide a name for the connection. This name identifies the vector store connection in ZBrain Builder and helps distinguish between multiple configured stores.

API Key

Enter the API key required to authenticate with the selected vector storage provider. This key is securely stored and used by ZBrain Builder to establish and maintain the connection.

An option is provided to show or hide the API key while entering it.

Environment

Specify the environment associated with the vector store (for example, a region or deployment environment). This value ensures that ZBrain Builder connects to the correct instance of the vector storage service.

Index Name

Enter the name of the index where vector data will be stored. This index is used to store embeddings generated during content ingestion and indexing.

The index name must match an existing index or be created according to the vector store’s configuration requirements.

Completing the connection

To save the configuration:

• Click Add to validate and create the vector store connection

• Click Cancel to discard the configuration and close the dialog

Once added, the vector store connection becomes available for use during data source configuration and knowledge base creation.

These settings determine how ZSearch processes, stores, and retrieves content from the data source.

Save the configuration to create the knowledge base.

Knowledge base creation

Once a connector is saved, ZBrain Builder automatically creates a dedicated Knowledge Base (KB) for that data source. This knowledge base serves as the structured, searchable representation of the connected content.

Key behaviors include:

• A single knowledge base is created per connector by default

• All content ingested from the data source is indexed into this knowledge base

• The knowledge base maintains a direct association with its source system

If Knowledge Graph is enabled during configuration, ZBrain Builder creates two knowledge bases for the same connector:

• One knowledge base optimized for the selected vector-based retrieval

• One knowledge base optimized for Knowledge Graph–based relationships and reasoning

This dual–knowledge base setup enables advanced semantic and relational search capabilities while preserving the same underlying data source.

Content and permission ingestion

ZSearch ingests both content and permission metadata from connected data sources. This ensures that search results and AI-generated responses respect the original access controls defined in the source system.

During ingestion:

• Files and documents are fetched from the connected repository

• User and group permissions associated with each file are captured

• Access Control Lists (ACLs) are preserved and enforced during search

Users can only see search results and project content they are authorized to access.

Continuous synchronization and indexing

Connected data sources are synchronized continuously. ZSearch recursively scans the configured source and automatically detects changes, including:

• Newly added files

• Updated content

• Newly created subfolders

Synchronization occurs at regular intervals (approximately every five minutes). Any new or updated content is automatically indexed and made available for search without requiring manual intervention.

This continuous synchronization ensures that enterprise search results remain current and aligned with the latest state of organizational knowledge.

Outcome of adding a data source

After a data source is successfully added and indexed:

• A searchable knowledge base is available to ZSearch

• Content is continuously kept up to date

• Permissions are enforced consistently across search and projects

• The data source becomes part of the enterprise context used by ZSearch

Once configured, the connected data source can be used immediately for enterprise search and project-based workflows.