Connectors

What are Connectors?

Connectors are reusable integration components that bridge external systems with the YuzeData platform. They provide a standardized way to connect to databases, REST APIs, cloud services, IoT platforms, and other data sources.

Templates and Instances

Connectors have two forms:

Concept	Description
Connector Template	The blueprint - defines what settings are needed and what operations are available
Connector Instance	A configured deployment with actual credentials for your organization

Connector templates available to your organization are configured based on your domain and use cases. When you deploy a connector, you're creating an instance from a template. Multiple instances can be created from the same template (e.g., connecting to different Salesforce orgs).

See HTTP API Connector for step-by-step instructions on setting up an HTTP API connector.

Operations

Each connector provides one or more operations - discrete actions it can perform against the external system.

What is an Operation?

An operation represents a specific capability of a connector, such as:

• Fetching records from an API
• Creating or updating data in an external system
• Triggering an action in a third-party service
• Synchronizing data between systems

Input and Output Schemas

Operations define schemas for their inputs and outputs:

• Input Schema: Specifies what data the operation needs to run (e.g., a date range, filter criteria, or record ID)
• Output Schema: Describes the structure of data returned by the operation

These schemas enable:

• Validation of data before execution
• Mapping between workflow data and connector data
• Clear contracts between connectors and workflows

Operation Capabilities

Each operation displays capability badges that tell you what advanced features it supports. These are detected automatically — you don't need to configure them.

Capability	What it means
Paging	Handles large datasets automatically by retrieving data in batches instead of all at once
Checkpoint resume	Remembers where it left off between runs, so subsequent executions only fetch new or changed data
Master data sync	Designed to synchronize master data with the platform
Dynamic output	Works with any data structure, not just a fixed format. Typically requires a schema mapping to transform the output into a known schema
Configurable	Has additional settings you can customize when deploying the operation

TIP

Capabilities help you pick the right operation for your use case. For example, if you need to sync large datasets incrementally, look for operations with both Paging and Checkpoint resume badges.

Operation Dependencies

Some operations depend on other operations to function. For example, an operation that processes orders might first need to fetch customer data.

Dependencies can be:

• Within the same connector: One operation calls another
• Cross-connector: An operation requires data from a different connector

Settings

Connectors are configured through settings - fields that define how to connect to the external system.

Setting Types

Type	Description	Example
Connection	URLs, endpoints, environment selection	API base URL
Authentication	Credentials for accessing the system	API key, username/password, OAuth tokens
Configuration	Behavioral options	Polling interval, batch size

Secret Settings

Sensitive settings (API keys, passwords, tokens) are marked as secrets:

• Encrypted at rest using customer-specific encryption keys
• Never displayed in plain text in the UI
• Securely passed to operations during execution

Invocation Modes

Connectors can execute in different ways depending on your needs:

Mode	Description	Use Case
Direct	Can be used directly in workflow steps	Workflow-driven data fetching
HTTP API	Exposed as an HTTP endpoint	Webhooks and external system callbacks
Batched	Collects requests and processes them together	High-volume operations
Scheduled	Runs automatically on a cron schedule	Periodic data synchronization

Direct Mode

In direct mode, workflows invoke connector operations to fetch or send data to external systems.

How it works:

1. A workflow step invokes a connector operation
2. The connector calls the external system (API, database, etc.)
3. The external system returns a response
4. The connector passes the data back to the workflow
5. The workflow can then produce datapoints, update master data, or continue processing

This is the most common pattern for integrations where YuzeData needs to pull data from external systems on demand.

HTTP API Mode

HTTP API mode exposes a connector as an inbound HTTP endpoint that external systems can call directly. Instead of YuzeData pulling data from an external system, the external system pushes data to YuzeData.

How it works:

1. External system sends an HTTP request to the connector's endpoint
2. YuzeData validates authentication (OAuth token or webhook signature)
3. The connector processes the request
4. Data can be written to datapoints, trigger workflows, or return a response

Authentication:

You can create API clients for fine-grained authorization control. Each API client can be granted access to specific connectors, allowing you to:

• Give different external systems their own credentials
• Control which connectors each system can access
• Revoke access independently per client

API clients are managed in Manage → API Clients.

Use cases:

• Receiving webhook events from external systems
• Accepting callbacks when external processes complete
• Allowing third-party systems to push data into YuzeData