Overview

A Flow is an organized set of Components which interact with each other via a simple and intuitive drag and drop interface to perform a specific task.


Flow Components

Flows differ in complexity depending on the purpose of the Flow. Components within a Flow are connected together by lines which represent data flow paths between the components.

When a Flow is run, components defined within the Flow are executed in order to produce one or more outputs as defined by the particular Flow.

If you click on the triple-dot icon at the top right of the trigger component you can change the name "Trigger" and you can write a description as you can see on the top right of the screen. This method works for every components.

In the illustration down below you can see all the different inputs that can be created. You can add components with the "+ add" button on the config. When you click on preview you can fulfill all your different inputs.

Each component output has a unique identifier which is created at the same time the instance of the component is created. The identifier of the trigger are "trigger" + "Variable Name". For example if you have a text the identifier is trigger.text and you can call your identifier by taping "@".


Component Data Flow

Data is transferred through a Flow with connections between components. For example, the Flow shown below, is a simple Flow, in which an Input component is used to feed a textual query (via a Text Input Component) into an OpenAI component. The OpenAI component has an output connected to an Output component, in order to receive the result of the query.

Components accept only certain types of data flow connections from other components. For example, the input of an OpenAI component will only accept a connection to a Text Input component, as the OpenAI component expects a textual input query.

A Flow can also be configured as an API end-point, thereby providing optimal adaptability for integration within your software application.

Components can be selected from a large library of pre-defined components and components can be created to encapsulate customized data processing tasks.


Component Parameters

Most components have configuration parameters. Depending on the particular component, key configuration parameters regarding the execution of the component can be selected directly in the component UI block.

For example, for the OpenAI component, we can select the particular OpenAI model to use by clicking the "Model" dropdown menu as shown below.

For more advanced configuration of a component, select the "Show Advanced configurations" at the bottom of the component and all the different configurations will appear providing options for component as shown below.

For more information regarding the available components and detailed information regarding each components configuration parameters, please refer to the Component Reference section.

Last updated