Use ProcessMaker's FlowGenie to generate AI-powered tasks by simply entering a description.
ProcessMaker's FlowGenie introduces an AI Studio within the Process Modeler, providing Process Designers with the capability to execute a wide range of tasks by simply entering a description of what they wish to achieve. FlowGenie has the following features and functionalities:
FlowGenie is accessible as a new AI-powered process object within the Modeler.
Designers enter a prompt and provide relevant input data, which the FlowGenie utilizes to generate a response.
The response can be formatted as text, JSON, or an HTML table.
Designers have the ability to add and link multiple Genies within their processes. Each Genie executes a single prompt and returns a response in a request variable for the next Genie. Request variables can be passed to Genies using mustache syntax within the prompt.
Genies can be expanded into a full-size FlowGenie Studio, enabling users to test, preview, and fine-tune their behavior directly within the Modeler environment.
Genies also have the capability to interpret file attachments, providing potential for document processing.
Transform existing case data: Merge string fields, perform calculations on number or date fields, and sort arrays.
Generate new case data: Create lists/tables, pre-fill responses, and translate inputs.
Analyze case data: Summarize texts and offer insights from common public information.
Analyze file attachments: Identify documents, scan images for context, and perform OCR extractions.
Watch the following product tour to learn how FlowGenie can analyze images and generate text output based on the entered prompt.
Watch the following product tour to learn how FlowGenie can use Request Variables to generate a JSON response.
Watch the following product tour to learn how FlowGenie can use a variety of input data and generate a response formatted as a table.
Add a FlowGenie object from one of the following locations in Process Modeler:
Object Panel: Located to the left of the Process Modeler canvas, the Object Panel contains various process model objects.
Object Bar: Located at the bottom of the Process Modeler canvas, the Object Bar contains pinned Process model objects, which are easy accessible.
Follow these steps to add a FlowGenie from the Object panel to the Process model:
Click the FlowGenie object from one of the following sections:
Pinned Objects: The FlowGenie object appears in the Pinned Objects section when it is pinned.
Click the location in the Process model to place this element. Follow these guidelines when placing this element:
Optionally place this element between two existing Process model objects already connected with a Sequence Flow element. See Place an Object Between Two Process Model Objects Already Connected with a Sequence Flow Element.
Follow these steps to add a FlowGenie object from the Object bar to the Process model:
Ensure that the object is pinned to the Object bar. If not, pin it.
In the Object bar at the bottom center, click the object's icon. This object's icon displays adjacent to the mouse icon to indicate this FlowGenie may be placed into the Process model.
Click the location in the Process model to place this object. Follow these guidelines when placing this element:
Optionally place this element between two existing Process model objects already connected with a Sequence Flow element. See Place an Object Between Two Process Model Objects Already Connected with a Sequence Flow Element.
Follow these steps to configure a FlowGenie:
In the Genie Name setting, enter a name for this object.
In the Variable Mapping section, configure the input and output setting for the FlowGenie.
In the Genie Output section, select a Response Format for the Genie's response. Choose from one of these options:
Text: Use this option to get the response as plain text.
JSON: Use this option to get the response formatted as a JSON object.
Table: Use this option to get the response formatted as an HTML table.
Click Preview to access the FlowGenie Studio. Here, you can test and preview the Genie's response, as well as adjust the configuration as necessary.
The FlowGenie Studio provides an interface to preview, test and adjust Genie configurations. Follow these steps to use the Studio:
In the Prompt setting, enter a new prompt or edit the current one.
From the Response Format setting, select one of the available formats.
Text: Use this option to get the response as plain text.
JSON: Use this option to get the response formatted as a JSON object.
Table: Use this option to get the response formatted as an HTML table.
In the Test Data setting, input data in JSON format for the Genie to utilize in executing the task outlined in the prompt.
Select the Include File settings to upload a file for processing, and use the Browse button to upload the file.
Click Run to test the FlowGenie. The response generated by the Genie will be displayed in the Response section.
Click Ok to save the FlowGenie configuration, otherwise click Cancel.
Ensure that the Object panel is visible. If not, click the Add icon from the Object bar at the bottom.
Object Category: The FlowGenie object appears in the Object Category section when it is not pinned.
If a Pool element is in your Process model, the FlowGenie element cannot be placed outside of the Pool element.
If a Pool element is in your Process model, the FlowGenie element cannot be placed outside of the Pool element.
Select the object from the Process Model, then click the Show Configuration button in the upper right corner of the Modeler to view the Configuration panel.
In the Prompt setting, enter a prompt for the FlowGenie to use. Case data can be used in prompts through mustache syntax.
In the Genie Input section, optionally select the Input File option to provide a file for the Genie to process.
In the Output Variable Name setting, use mustache syntax to specify a request variable for storing the response returned by the Genie. This variable can then be referenced elsewhere in the process to utilize the Genie's response.
Administrators have the flexibility to customize FlowGenie configuration to suit specific requirements and adjust the generated responses as needed.
Follow these steps to view and edit FlowGenie configurations:
Ensure that you are logged on to ProcessMaker.
Click the Admin option from the top menu. The Users page will display.
From the Settings panel on the left, expand the Integrations section, and then select AI settings.
Text Model: This setting is used to determine the OpenAI model utilized when no file is included as part of the input.
Vision Model: This setting specifies the OpenAI model used when an image is provided as input.
Text System Prompt: This setting defines the system prompt utilized when no file is included as part of the input. Refer to OpenAI's guide for more details.
Vision System Prompt: This setting defines the system prompt utilized when an image is provided as part of the input. Refer to OpenAI's guide for more details.
Text JSON Prompt Suffix: This setting is used to append text to the Text System Prompt when the response format is set to JSON.
Vision JSON Prompt Suffix: This setting is used to append text to the Vision System Prompt when the response format is set to JSON.
Table Prompt Suffix: This setting is used to append text to the system prompt when the response format is set to Table.
Text Model Temperature: This setting specifies the temperature parameter used with the Vision Model. Refer to OpenAI's guide for more details.
Vision Model Temperature: This setting specifies the temperature parameter used with the Text Model. Refer to OpenAI's guide for more details.
Click the Settings icon from the left sidebar to view all settings.
Use the Edit icon to change the value of a setting.
Use the Copy icon to copy the value of a setting.
Use the Clear icon to clear the value of setting.