Skip to main content
All posts
Product7 min read

Connecting 17+ Data Sources: How Enterprise Search Connectors Work

From Confluence to Salesforce to PostgreSQL — ZenSearch connects to 17+ data sources with authentication, incremental sync, permission import, and rate limiting.

December 30, 2025 · ZenSearch Team

ZenSearch connects to 17+ data sources through one unified connector framework: ten document and collaboration platforms (Confluence, Slack, GitHub, Jira, Notion, Google Drive, SharePoint, S3, Azure Blob, Web Crawler), three CRM and ERP systems (Salesforce, HubSpot, SAP), and four databases accessed via natural-language-to-SQL (PostgreSQL, MySQL, ClickHouse, SQL Server). Every connector handles authentication, incremental sync, and permission import through the same lifecycle, so adding a new source doesn't require bespoke plumbing.

Enterprise knowledge doesn't live in one place. It's distributed across documentation platforms, code repositories, messaging tools, cloud storage, CRM systems, and databases. ZenSearch connects to 17+ data sources through a unified connector framework.

How Connectors Work

Every ZenSearch connector follows the same lifecycle:

1. Authentication — Each connector supports its source system's native auth methods. Confluence uses API tokens. Google Drive supports both OAuth and Service Account. Salesforce offers OAuth and password auth. Database connectors support SSL/TLS and platform-native authentication.

2. Discovery — The connector enumerates available content: Confluence spaces, SharePoint sites, GitHub repos, Slack channels, Notion workspaces, Salesforce objects, and database schemas.

3. Collection — Documents are fetched with their metadata and permissions. Connectors respect rate limits, handle pagination, and manage API quotas automatically.

4. Incremental Sync — After the initial full sync, connectors track changes using source-specific mechanisms (timestamps, change tokens, webhooks). Only new and modified content is re-processed, keeping your index current without redundant work.

5. Permission Import — Access control lists are extracted and stored alongside document content for query-time filtering.

Document & Collaboration Connectors

ConnectorKey Features
ConfluenceCloud & Server, spaces, attachments, permission sync
NotionPages, databases, nested blocks, full block-level indexing
Google DriveOAuth + Service Account, format conversion (Docs/Sheets/Slides)
SharePointMulti-site, version history, permission inheritance
GitHubRepos, PRs, issues, code search, branch selection, Enterprise
JiraJQL filtering, comments, worklogs, attachments
SlackChannels, threads, files, reactions, date range filtering

Storage & Web

ConnectorKey Features
Amazon S3IAM roles, S3-compatible storage, prefix filtering
Azure BlobMultiple auth methods, container filtering, SAS tokens
Web CrawlerDepth control, robots.txt compliance, rate limiting

CRM & ERP

ConnectorKey Features
SalesforceOAuth, advanced queries, bulk operations, sharing permissions
HubSpotContacts, Companies, Deals, Tickets, engagement history
SAPOData integration, Business Partner, Sales Orders

Database Connectors

ConnectorKey Features
PostgreSQLSchema discovery, natural language queries, SSL
MySQLTLS, charset config, table and view filtering
ClickHouseOptimized for analytics workloads
SQL ServerWindows and SQL auth, encrypted connections

Database connectors work differently from document connectors. Instead of indexing content, they enable natural language queries that are translated to SQL and executed in read-only mode against your database. Schema metadata is indexed to help the AI understand your data model.

Custom Connectors

Enterprise customers can request custom connectors for proprietary systems. The connector framework handles authentication, rate limiting, pagination, and error handling — custom connectors only need to implement the source-specific logic.