Skip to main content
This guide walks you through the steps to get your first sub-account set up, create a data source connection, and start querying data using AI. Before you start, it helps to understand a few key terms used throughout the guide.

​
Key Features

  • Broad Data Connectivity: Connect AI Embed allows you to transform data into a simple language that any reporting tool, analytics dashboard, or spreadsheet can understand. It can connect to a vast range of applications (including those for accounting, marketing, customer relationship management, and more) by using a consistent approach that is already supported by many data tools.
  • Live Data Access: Connect AI Embed gives you live access to your data—queries run directly against the source system at the moment they are made, without moving data to a separate location first. This approach avoids the pitfalls of traditional Extract, Transform, Load (ETL) processes that rely on moving data to separate locations.
  • Read and Write Support: Connect AI Embed supports both read and write operations across hundreds of integrated data sources, allowing AI agents and embedded applications to query, create, update, and delete records. Operations are controlled by security settings that you can configure in Connect AI Embed or in your data sources.
  • Secure Access: Administrators can create query users and define their access permissions for each data source individually.
  • Comprehensive Data Model & Dynamic Discovery: Connect AI Embed provides comprehensive access to all of the data that is exposed in the underlying connections, including full access to dynamic data and searchable metadata.
For more information about these key features, see Technology.

​
Key Terms

  • Parent Account—Your top-level CData account. As the Independent Software Vendor (ISV), this is the account you use to manage your entire Connect AI Embed deployment, including creating and overseeing all sub-accounts.
  • Sub-Account—An account that sits beneath your parent account, representing one of your customers or tenants. Sub-accounts are where connections are created and data is queried. Each sub-account is isolated from the others.
  • ISV/Embedded Customer—These terms all refer to the same entity: your company, which is building a product on top of Connect AI Embed. You control the experience your end users see.
  • End User—An individual user within one of your customers’ sub-accounts. End users interact with your application. They do not interact directly with Connect AI Embed except during initial data source setup.
  • Connection—A configured link to a data source (for example, Salesforce, Jira, or ServiceNow). Connections belong to a sub-account and can be shared across all end users in that sub-account, or set up individually per end user.
  • MCP (Model Context Protocol)—An open standard that allows AI agents to connect to and query data sources. Connect AI Embed exposes MCP servers that your AI agents use to access connected data. The standard endpoint provides access across all connected sources. Connect AI Embed administrators can also scope MCP access to specific connections, limiting the end users’ data access. See MCP in the API Reference for details.
  • JWT (JSON Web Token)—A secure, signed token used to authenticate all API requests to Connect AI Embed. Every JWT must identify the parent account (iss). Most calls also require the sub-account ID (sub). For scoped MCP server access, the JWT must additionally include a connection_ids claim specifying which connections the end user can access. See Authentication for details.

​
Get Started

For a high-level overview about setting up and using Connect AI Embed, see Quick Start Guide. For more detailed information about the features of Connect AI Embed, see the documentation outlined below:
  • Navigation—Information about navigating Connect AI Embed and accessing features.
  • Dashboard—An overview of recent queries and account activity for a selected sub-account, as well as a summary of all sub-accounts.
  • Accounts—An overview of the names, account Ids, and number of connections of each connected sub-account.
  • Connection API—The API to connect data sources to Connect AI Embed.
  • SQL Reference—List of supported SQL syntax and code examples.
  • Reference Architecture—Diagram of the three zones: your infrastructure, the Connect AI Embed managed service, and your customers’ data sources.
  • Connect Gateway—Live, read-write access to on-premises data sources without moving data outside a firewall.
  • Build Your First Agent—A developer-oriented guide: create a sub-account, connect a source, and start querying live data through MCP.

​
Get Support

For assistance with using Connect AI Embed, there are two main resources available:
  • The Status page, which shows the current status and incident history for the application services.
  • CData Technical Support, which you can also access by clicking the Support link at the bottom of the dashboard.