Tools in VIDIZMO Agentic AI
Tools extend the capabilities of VIDIZMO Agentic AI workflows by enabling AI nodes to access external information, systems, and specialized functionality. Tools are designed to support agentic behavior by allowing AI models to retrieve data, perform searches, and interact with services dynamically during workflow execution.
In the Workflow Designer, tools are added as nodes and connected to AI nodes. Once connected, tools become available to the AI model and can be invoked as needed to support response generation and decision-making.
What are Tools?
Tools are specialized nodes that expose focused, task-specific capabilities to AI models during workflow execution. They handle operations that fall outside the core reasoning capabilities of large language models, such as searching content repositories or retrieving current information.
Tools do not execute sequentially like standard workflow nodes. Instead, they are invoked by AI nodes when the model determines additional information or external interaction is required.
Purpose of Tools in workflows
Tools enable workflows to:
-
Access external or system-managed data sources
-
Retrieve current or dynamically changing information
-
Support AI responses with verifiable content
-
Extend AI behavior without increasing workflow complexity
-
Enable autonomous, agent-driven decision-making
By separating tool-based actions from workflow logic, workflows remain modular, adaptable, and easier to maintain over time.
Tools and AI nodes
Tools are designed to work in conjunction with AI nodes, such as Large Language Model nodes. When connected, the AI node can reference available tools during execution and decide whether to use them based on user input and workflow context.
This interaction allows:
-
AI nodes to focus on reasoning and language generation
-
Tools to handle deterministic, external, or structured operations
-
Workflows to remain flexible without enforcing rigid execution paths
Tool nodes and regular nodes
Tools differ from regular workflow nodes in both purpose and execution model.
| Aspect | Tool nodes | Regular nodes |
|---|---|---|
| Execution | Invoked by AI when needed | Execute sequentially |
| Control | AI-driven | Workflow-defined |
| Purpose | Extend AI capabilities | Perform workflow operations |
| Connection | Connected directly to AI nodes | Connected in execution flow |
When to use tools
Use tools when workflows require:
-
AI-driven decisions about data retrieval
-
Access to external or dynamic information
-
Responses grounded in organizational or real-time content
-
Flexible, adaptive AI behavior
Use regular nodes when execution must always occur in a fixed sequence or when deterministic control over processing is required.
Best practices for using tools
-
Connect tools only to AI nodes that require them
-
Keep tool responsibilities focused and well-defined
-
Limit tool access to relevant data and systems
-
Document tool usage and intent within the workflow
-
Review tool configurations as workflows evolve
Summary
Tools enhance workflows by providing targeted functionality that complements AI reasoning. They enable workflows to search content, retrieve external data, and interact with systems beyond the workflow itself. By using tools effectively, you can build scalable, maintainable, and intelligent workflows that respond accurately to user needs.