Agentic Workflow
An Agentic Workflow is a low-code tool for building and running automated, multi-step processes. Use it to design workflows, assign tasks to AI agents, and monitor execution.
To access this page, go to the home page > Agents > Agentic Workflows.
The Agentic Workflows page opens and displays a list of all previously created workflow journeys, if available. This allows you to quickly review, manage, or extend existing workflow configurations.
Use the Agentic Workflow environment to design and activate workflows. Once activated, they can be executed by multiple agents, triggered by specific events, or managed while interacting with agents. As they run, they can be monitored for performance, patterns, and anomalies, with insights feeding back into workflow optimization, making the entire automation cycle adaptive and self-improving.
 |
Creating an App
You can start a new workflow journey by taking the first step, creating an App. An App is like a container that holds all your Workflow configurations in one place.
Note
The platform supports multi-user access, allowing multiple users to work on an app and its journeys simultaneously.
Go to the Home page > Agents > Agentic Workflow > Start Your Journey.
The Apps page opens and displays a list of existing apps, if available.
If there are no applications yet, click BUILD APPS. If apps already exist, you will see a list of applications. Click CREATE APP in the top-right corner.
In either case, the Create App window opens.
Enter a preferred name for your app in the Display Name field.
The Name (identifier) field is auto-filled; however, you can update it as needed.
Note
The system uses this name internally. Ensure it is unique, concise, and follows the recommended naming conventions.
Use the A8Flow App Type (selected by default). The other app types are not currently available.
Set the Avatar for your app, if needed.
Click Upload.
Select an image from your local system.
Click Open.
The selected image is uploaded and set as the new avatar for your app.
Click CREATE.
A success message confirms the app was created.
Managing an App
You can manage all your existing apps from the Apps page, where you can easily edit or remove them as needed.
Editing an App
To update an app:
Click the Edit
icon next to the required app.The EDIT APP window opens.
Update the app name and avatar as needed.
Note
Name (identifier) and App Type fields are non-editable.
Click SAVE.
A success message confirms the app was updated.
Removing an App
To remove an app:
Click the Delete
icon next to the required app.The DELETE APP confirmation box opens.
Re-enter the name of the app you want to delete for confirmation.
Click Delete.
A success message confirms the app was removed from the list.
Creating Journeys
A Journey defines your application's entire workflow, guiding it from start to finish. It outlines each step, decision, and action, ensuring that every task is executed in the correct sequence and the overall process runs smoothly. By mapping your workflow as a journey, you gain clarity, control, and efficiency in managing your operations.
Within an app, you can create multiple journeys to organize and manage different workflows in one place.
To create a new journey:
Click on the desired app.
The Assisted page opens, and the left menu Journey is selected by default.
Click GET STARTED or CREATE JOURNEY. (If there are no other journeys present, you will see the GET STARTED button.)
The Create Journey window opens.
Enter a name for the journey in the Assisted Name field.
Enter a Description of your workflow journey.
Click CREATE.
A success message confirms the journey was created. You can find the journey on the Agentic Workflows page.
You can repeat these steps to create multiple journeys within the same app. Assisted journeys are ideal for business processes that involve multiple participants and require coordinated teamwork.
Note
Because workflow creation depends on the required journeys being available, you should create all the required journeys for the app before starting workflow creation.
Managing Journeys
You can manage all your existing journeys in the app, easily editing or removing them as needed.
Editing a Journey
To update a journey:
Click the Edit
icon next to the required journey.The EDIT JOURNEY window opens.
Update the name and description as needed.
Click SAVE.
A success message confirms the journey was updated.
Removing a Journey
To remove a journey:
Click the Delete
icon next to the required journey.The DELETE JOURNEY confirmation box opens.
Re-enter the name of the journey you want to delete.
Click Delete.
A success message confirms the journey was deleted.
Creating a Workflow Diagram
A workflow is a set of interdependent tasks or processes that work in tandem to achieve a common business goal. This series of tasks assigned within a workflow has preset, user-defined rules and conditions. The processes are linear and can be completed sequentially or in parallel.
After creating an app and its journeys, you can start designing the workflow for each journey.
Tip
You can design your workflow using a diagram based on the Business Process Model and Notation (BPMN) standard. Refer to BPMN Framework for more information.
To create a new workflow:
Click on the desired app and its journey.
The journey page opens with a plain canvas.
Start building your workflow process and configure it as needed.
Manual Workflow - Create workflows manually using the tools available on the page. Drag and drop tasks, connect them with arrows as needed. Refer to Creating Workflow Tasks for more information.
Automatic Workflow - The platform includes an AI assistant that can automatically generate workflows from your natural-language descriptions.
Click the AI Assistant
icon to open the AI assistant at the bottom of the page.Enter your workflow requirements in the text box as a prompt.
Press Enter.
The AI assistant analyzes your prompt and automatically constructs a complete workflow diagram, including relevant task types, sequence flows, and integration of existing AI agents from the Agent SDK, as well as MCP tools from the Agent Action Library for consistency.
If the generated workflow does not match your requirements, you can edit it manually or re-enter a more detailed prompt.
Tip
You can also use the AI assistant to modify existing workflows.
After completing the workflow creation, configure each element as needed. Refer to Creating Workflow Tasks for more information.
Click SAVE to save your changes.
Saving stores your current changes to the canvas, preserving your work and keeping it available for further editing.
A success message confirms your changes were saved.
When multiple users are collaborating on the same journey, you must click COMMIT to save your changes and commit the canvas for reuse.
The Commit confirmation box opens.
Note
This action creates a new version on the canvas, ensuring proper version control.
Click SAVE & COMMIT to save the changes and commit the canvas for reuse.
A success message is displayed, and the canvas becomes read-only.
Click DISCARD & COMMIT to discard the changes and still commit the canvas for reuse.
A success message is displayed, and the canvas becomes read-only.
Creating Workflow Tasks
The Workflow Journey Canvas is your visual design environment for creating workflow processes. Built on BPMN standards, it enables you to design workflows through an intuitive drag-and-drop interface.
By using BPMN, you can clearly define the steps, decisions, and agents involved in a process, making it easier to build, debug, and optimize. It ensures that both humans and AI systems have a shared understanding of how a task should be accomplished.
The journey page is organized into three distinct sections, each serving a specific purpose in the workflow design process:
Left Panel - Contains all the building blocks you need to construct your workflow diagram, including various task types, decision points, and connecting elements.
Center Canvas - Your main workspace where you assemble and arrange workflow elements to visualize your process flow.
Right Panel - A configuration panel that provides access to settings and properties. Use this panel to customize the behavior and parameters of selected workflow elements.
 |
While creating workflow tasks, make sure to follow these basic BPMN standard rules:
Every workflow must have a Start Event and End Event.
All tasks must be connected with sequence flows.
Each task requires at least one incoming and one outgoing connection.
Gateways (decision points that control process flow) must have one incoming flow and multiple outgoing flows (or vice versa).
No orphaned or disconnected elements are allowed.
Flows must follow a logical direction from start to end.
Tip
The platform supports two methods for creating workflow diagrams:
Pop-up menu method
Drag-and-drop method
To create a simple diagram:
The Start Event is available on the canvas by default.
Click Start Event.
The pop-up menu opens.
Click the Task
icon.The blank rectangular task shape is added to the canvas.
Click the task and click the Change Type
icon from the pop-up menu.The list of all available task types is displayed. Each task type represents a different set of actions and behaviors that the task can perform within your workflow. Refer to Tasks Configuration for detailed information.
Click Service Task from the list.
The task on the canvas is highlighted and displays a small icon, indicating the task type you've chosen.
Click the task, then in the right-side configuration panel, under the General tab, fill in the ID and Name for the task. Refer to Tasks Configuration for detailed information.
Additionally, you can double-click on the task.
The Service Tasks window opens.
Choose any existing service and click Select to execute. Refer to Tasks Configuration for more information.
Click SAVE on the workflow canvas to save the changes.
Drag the aiAgent Task shape from the left-side panel directly onto the canvas.
Double-click on the aiAgent task.
The AI Agents window opens and displays a list of public AI Agents that you created in the platform.
Search for the required AI agent and select it from the filtered list.
Click Submit to tag the AI Agent in the task.
Click Service Task.
The pop-up menu opens.
Click the Connector
icon and connect it to the AI Agent Task.
You can add an Action Task and tag an MCP tool that you created on the platform.
Similarly, you can add Multiple Tasks as needed and Configure them.
After creating all tasks, click your final task and select End Event from the pop-up menu to complete your workflow.
The End Event shape is added to the canvas.
Click SAVE to save the canvas.
Task Types
The platform offers a variety of BPMN task types, enabling quick workflow creation and straightforward configuration. Each task type is designed to handle specific actions within your workflow, making it easy to build complex processes.
Action Tasks
Action Tasks represent workflow activities that invoke MCP tools that you created in the Agent Action Library. Action tasks bring the power of MCP integrations directly into your workflows, enabling your processes to interact with external systems, access real-time data, and perform operations across connected services.
When a workflow reaches an action task, the system automatically invokes the MCP tool and sends any required input parameters. The MCP tool connects to the external systems or database, performs the requested operation, and returns the results. Once the MCP tool returns its response, the workflow engine captures the data and continues execution to the next step in the process.
Common uses for action tasks include:
Search for customer information in your CRM system
Query databases and fetch relevant records
Retrieve and process files from cloud storage
Create or update tasks in project management tools
Fetch real-time data from external APIs
AI Agent Tasks
AI Agent Tasks represent workflow activities that invoke AI agents that you created in the Agent SDK. AI Agent tasks bring AI capabilities directly into your workflows, enabling your processes to leverage custom-built agents for intelligent automation, data processing, and decision-making.
When a workflow reaches an AI agent task, the system automatically invokes the AI agent and sends the input query or data. The AI agent processes the information using its trained capabilities, analyzing content, extracting insights, generating outputs, and making intelligent decisions. Once the AI agent returns its response, the workflow engine captures the results and continues execution to the next step in the process.
Common uses for AI agent tasks include:
Analyze customer feedback and sentiment
Extract structured data from unstructured documents
Generate reports or summaries automatically
Classify and route information intelligently
Provide context-aware recommendations
Business Rule Tasks
Business Rule Tasks enable you to evaluate business rules within your workflows. These rules are defined using Decision Model and Notation (DMN), a standard language for modeling and executing business decisions in a structured, consistent manner.
When a workflow reaches a business rule task, the system evaluates the defined business rules and stores the result in a process variable. The workflow then continues execution based on the evaluation outcome and moves to the next step in the process.
Common uses for business rule tasks include:
Approving loan applications
Determining discounts for customer orders
Routing customer service requests to specific teams
Calculating risk levels for new business ventures
Manual Tasks
A Manual Task is a special kind of task, used to model a task that is performed by a human actor outside of the BPM engine. Manual tasks are typically used to model tasks that are difficult or impossible to automate, such as:
Approving a document
Making a decision
Completing a physical task
For example, if a process contains a step for "Verify physical documents", that step can be modeled as a manual task because the actual work happens outside the system.
When a process instance reaches a manual task, the engine treats it as a pass-through activity; the process automatically continues to the next step when execution reaches it.
The platform supports manual tasks for modeling and process readability, but they do not behave like other types of tasks, where execution waits for interaction or a system response.
Receive Tasks
Receive Tasks represent workflow activities that pause workflow execution until a specific message or event is received from an external system. Receive tasks create controlled wait states in which your automated processes must pause until incoming communication arrives.
When a workflow reaches a receive task, the process instance stops at that point and enters a waiting state until the expected message arrives from an external source. Once the message is received and matched to the waiting process instance, the workflow captures the message data and continues execution to the next step in the process.
Common uses for receive tasks include:
Waiting for customer approval responses before proceeding with order fulfillment
Pausing until payment confirmation is received from a payment gateway
Holding for data synchronization signals from external systems
Waiting for partner system acknowledgments
Stopping until document upload confirmations are received from users
Script Tasks
Script Tasks represent workflow activities that execute custom scripts or code within your workflows, providing flexibility for specialized logic and data manipulation. Script tasks enable you to embed programming logic directly into your processes for business operations.
When a workflow reaches a script task, the system executes the configured script code in the specified programming language. The script can access and modify workflow variables, perform calculations, transform data formats, or implement custom business logic. Once the script completes execution, the workflow proceeds to the next step in the process.
Common uses for script tasks include:
Transforming data formats between different task requirements
Calculating values based on multiple workflow variables
Implementing specialized validation or business rules
Constructing dynamic data structures (JSON, XML) for external API calls
Manipulating collections or arrays of data within the process context
Important
Script tasks offer flexibility for custom logic, but they require careful testing and maintenance to ensure reliability and security.
Send Tasks
Send Tasks represent workflow activities that transmit messages or notifications to external systems as part of a business process. Send tasks enable outbound communication from workflows to external recipients, systems, or message brokers.
When a workflow reaches a send task, the system creates a work item for message transmission and executes the send operation through authorized workers. Once the message is successfully transmitted to the external system, the workflow continues execution to the next step in the process.
Common uses for send tasks include:
Sending email confirmations to customers
Publishing events to streaming platforms like Kafka to trigger downstream processes
Delivering status change notifications to external applications
Transmitting alerts to monitoring or logging systems
Service Tasks
Service Tasks represent workflow activities that invoke external services or systems as part of a business process. Service tasks are used for automated operations that interact with web services, APIs, external applications, or background workers.
When a workflow reaches a service task, the system automatically invokes the configured external service and sends any required data. Once the external service completes its operation and returns a response, the workflow engine captures the result and continues execution to the next step in the process.
Common uses for service tasks include:
Sending email notifications
Calling external APIs
Starting new sub-processes
Interacting with databases
User Tasks
User Tasks represent workflow activities that require human interaction within the workflow system. User tasks are used for activities that need human decision-making, data entry, approval, or review as part of an automated business process.
When a workflow reaches a user task, the platform creates a work item and assigns it to a specific user or group. The process instance pauses at that point, waiting for the assigned user to complete the task through the workflow interface. Once the user submits their response, the workflow engine resumes execution and continues to the next step in the process.
Common uses for user tasks include:
Approving purchase orders
Reviewing documents
Completing customer service requests
Performing quality inspections
Making business decisions
Webhook Tasks
Webhook Tasks represent workflow activities that interact with external systems requiring asynchronous callbacks before the workflow can continue. Webhook tasks enable two-way communication between your workflows and external platforms, making them ideal for processes that depend on external events or actions to proceed.
When a workflow reaches a webhook task, the system initiates the outbound request to the external system and triggers the external event. The process instance then pauses and enters a waiting state until the external system sends a callback response. Once the callback is received and processed, the workflow engine captures the result and continues execution to the next step in the process.
Common uses for webhook tasks include:
Sending documents to e-signature platforms and waiting for signed confirmations
Initiating payment requests and waiting for transaction completion callbacks
Triggering external verification processes and receiving approval results
Starting third-party background checks and capturing the returned findings
Requesting data processing from external systems and waiting for completion signals
Tasks Configuration
Once you've added tasks to your workflow canvas, the next step is to configure them to meet your specific requirements. Each task type offers a set of configuration options that control how the task behaves, what data it processes, and how it interacts with other workflow elements.
This section describes how to configure each task type, including essential properties, optional settings, and configuration options.
Action Tasks Configuration
An Action Task represents work performed by an MCP tool within a business process. Action tasks are typically used to model activities that interact with external systems, such as retrieving data, updating records, or executing operations, and connect with MCP tools configured in the Agent Action Library environment rather than requiring human interaction.
Action task attributes:
Tabs | Attribute | Description |
|---|---|---|
General | Name | The "Name" attribute defines the name or label of the action task. |
Parameters | Input Parameters > Name | The "Name" attribute specifies the input parameter's name as defined in the MCP tool. This corresponds to the parameter names defined in your MCP tool configuration. |
Input Parameters > Value | The "Value" attribute specifies the value for the MCP tool's input parameter. This can be a static value or a workflow variable. | |
Output Variable | The "Output Variable" attribute specifies the variable name that stores the Action Task's output value. Use this variable to reference the agent's output in subsequent workflow tasks. |
AI Agent Tasks Configuration
An AI Agent Task represents the work performed by an AI agent within a business process. AI agent tasks are typically used to model activities that require intelligent processing, such as analysis, content generation, or decision-making, and interact with MCP tools tagged with the AI agent rather than requiring human interaction.
AI agent task attributes:
Tabs | Attribute | Description |
|---|---|---|
General | Name | The "Name" attribute defines the name or label of the AI agent task. |
Input Query | The "Input Query" attribute defines the variable or string that serves as input for the AI Agent. This query provides the context or question that the AI agent will process. | |
Parameters | Output Parameters > Name | The "Name" attribute defines the variable name or label of the AI agent output value. |
Output Parameters > Value | The "Value" attribute defines the temporary variable name that stores the AI agent output value. Use this variable to reference the agent's output in subsequent workflow tasks. |
Business Rule Tasks Configuration
A Business Rule Task is a specific BPMN activity that evaluates and applies business rules or decision logic as part of the process flow.
Key Characteristics of a Business Rule Task
Rule Evaluation: Business rule tasks represent activities that involve the evaluation of business rules, decision tables, or decision logic to determine the path the process should follow.
Decision Making: Business rule tasks are used to make decisions or calculations based on the provided data and rules, which can affect the subsequent flow of the process.
Automated or Semi-Automated: Business rule tasks can be automated, where the decision logic is executed by a rule engine or a system, or can involve human interaction for rule evaluation and decision approval.
Conditional Flows: Business rule tasks often have multiple outgoing sequence flows with associated conditions that define different paths the process can take based on the outcomes of the rule evaluation.
Attributes
Business rule task attributes:
Attribute | Description |
|---|---|
ID: | The ID attribute represents the unique identifier of the business rule task. It corresponds to the task definition key in the BPMN model. |
Name: | The "Name" attribute defines the name or label of the business rule task. |
Implementation
The "Implementation" attribute specifies the method, class, or service that the business rule task will invoke or execute. It defines the technical implementation behind the task's functionality. Implementation options include:
Implementation | Description |
|---|---|
DMN: | Decision Model and Notation (DMN) is a visual representation of decision logic within a business process. When selecting this option, you need to specify the DMN decision reference. Decision Ref is a reference to a specific decision within the DMN model that the business rule task will execute. Binding: Binding determines how the decision is linked to the business rule task. Our tool offers the following options:
Tenant ID: Tenant ID isolates decision execution within a specific tenant's scope, which is useful for multi-tenancy scenarios. Result Variable: Result Variable is where the output of the executed decision is stored, enabling further processing within the same process. |
Asynchronous Continuations
Use asynchronous continuations to control whether a task runs asynchronously:
Attribute | Description |
|---|---|
Asynchronous Before: | The "Asynchronous Before" attribute indicates if the task can be started asynchronously before other tasks or events in the process. |
Asynchronous After: | Similarly, the "Asynchronous After" attribute denotes whether the business rule task can be completed asynchronously after other tasks or events in the process. |
Manual Tasks Configuration
Within BPMN (Business Process Model and Notation), a Manual Task represents work that requires human intervention, such as decision-making or manual interaction.
Key Characteristics of a Manual Task
Human Interaction: Manual tasks represent activities that involve human intervention, requiring a person to perform specific actions, such as approval, validation, or other manual tasks.
Time-Driven or Event-Driven: Manual tasks can be triggered by specific events or deadlines (event-driven) or simply be part of the process flow (time-driven).
Attributes
Manual task attributes:
Attribute | Description |
|---|---|
ID: | The "ID" attribute represents the unique identifier of the manual task. It corresponds to the task definition key in the BPMN model. |
Name: | The "Name" attribute defines the name or label of the manual task. |
Asynchronous Continuations
Use asynchronous continuations to control whether a task runs asynchronously:
Attribute | Description |
|---|---|
Asynchronous Before: | The "Asynchronous Before" attribute indicates if the task can be started asynchronously before other tasks or events in the process. |
Asynchronous After: | Similarly, the "Asynchronous After" attribute denotes whether the manual task can be completed asynchronously after other tasks or events in the process. |
Receive Tasks Configuration
In BPMN, a Receive Task is a type of task used to model a specific type of activity in a business process that represents the receipt or reception of a message, notification, or request from an external participant or system. Receive tasks are typically used to depict a point in the process where the process is waiting for an external event, message, or response.
Key Characteristics of a Receive Task
Message Reception: Receive tasks represent activities where the process is waiting to receive a message, notification, or request from an external participant or system.
Asynchronous: Receive tasks are typically asynchronous, meaning that the process waits for an external event or message to arrive and continues its flow after receiving it.
Message Content: Receive tasks may involve specifying the expected content or format of the incoming message and how the process should handle it.
Response Handling: Receive tasks are often followed by subsequent activities that process the received message or response.
Integration: Receive Tasks are used to model the integration of the business process with external participants, systems, or services through message-based communication.
Attributes
Receive task attributes:
Attribute | Description |
|---|---|
ID: | The "ID" attribute represents the unique identifier of the receive task. It corresponds to the task definition key in the BPMN model. |
Name: | The "Name" attribute defines the name or label of the receive task. |
Details
The "Details" section for the receive task includes the following Message attributes:
Attribute | Description |
|---|---|
Message: [Message_0vat9jk (id=Message_1ubmceq)] | The "Message" attribute indicates the specific message that is associated with this receive task. In this example, it references a message with the identifier "Message_0vat9jk" and the unique ID "Message_1ubmceq." The message defines the content and structure of the data that the Receive task will receive. |
Message Name: [Message_0vat9jk] | The "Message Name" attribute specifies the name or identifier of the message that is expected by the receive task. In this example, it is identified as "Message_0vat9jk." The message name is used to correlate the received message with the corresponding message definition in the BPMN process. |
Asynchronous Continuations
Use asynchronous continuations to control whether a task runs asynchronously:
Attribute | Description |
|---|---|
Asynchronous Before: | The "Asynchronous Before" attribute indicates if the task can be started asynchronously before other tasks or events in the process. |
Asynchronous After: | Similarly, the "Asynchronous After" attribute denotes whether the receive task can be completed asynchronously after other tasks or events in the process. |
Script Tasks Configuration
In BPMN, a Script Task is a type of task used to model a specific activity in a business process where a predefined script or code is executed. This script can be written in a scripting or programming language and performs a specific action or set of actions within the process. Script tasks are often used for tasks that involve custom logic, calculations, or operations that cannot be represented using other BPMN elements.
Key Characteristics of a Script Task
Script Execution: Script tasks represent activities where a script, often written in a scripting language like JavaScript or a programming language like Java or Python, is executed as part of the process.
Custom Logic: Script tasks are used for tasks that require custom or specialized logic, calculations, or operations that cannot be easily expressed using standard BPMN elements.
Automation: Script tasks can be automated, meaning they do not require human intervention, and the execution of the script is performed by a software system.
Input and Output Data: Script tasks may involve input data that is used by the script and produce output data based on the script's execution. BPMN provides elements for modeling data associations between tasks.
Integration: Script tasks are sometimes used to integrate the process with external systems or services by executing scripts that interact with these external entities.
Attributes
Script task attributes:
Attribute | Description |
|---|---|
ID: | The "ID" attribute represents the unique identifier of the script task. It corresponds to the task definition key in the BPMN model. |
Name: | The "Name" attribute defines the name or label of the script task. |
Details
The "Details" section for the script task includes the following Message attributes:
Attribute | Description |
|---|---|
Script Format: | Specifies the language or format of the script, influencing how the content is interpreted. |
Script Type: | Determines how the script is handled within a script task in BPMN, choosing between inline or external storage. - Inline Script: Directly provided within the BPMN process, defining actions or behavior. Script: Requires providing the script content. - External Resource: Script stored externally for reusability. Resource: Requires specifying the location or reference, such as a file path or URL. |
Result Variable: | Designates the variable name to store the script's execution result. Accessible in subsequent process steps to capture the outcome. |
Asynchronous Continuations
Use asynchronous continuations to control whether a task runs asynchronously:
Attribute | Description |
|---|---|
Asynchronous Before: | The "Asynchronous Before" attribute indicates if the task can be started asynchronously before other tasks or events in the process. |
Asynchronous After: | Similarly, the "Asynchronous After" attribute denotes whether the script task can be completed asynchronously after other tasks or events in the process. |
Send Tasks Configuration
A Send Task models a specific type of activity that represents the process of sending a message, notification, or request to an external participant or system. Send tasks are used to depict a point in the process where information is transmitted to another entity, such as a human user, another process, a service, or an external system.
Key Characteristics of a Send Task
Message Sending: Send tasks represent activities where a message, request, or notification is sent to an external participant or system as part of the process.
Asynchronous: Send tasks are typically asynchronous, meaning that the process does not wait for an immediate response but continues its flow.
Message Content: Send tasks may involve specifying the content of the message to be sent, which can include data and details relevant to the interaction.
No Direct Response: Unlike "receive tasks," which represent waiting for a response or incoming message, send tasks do not expect a direct response from the external entity.
Integration: Send tasks are often used to model the integration of a business process with external participants, systems, or services through message-based communication.
Attributes
Send task attributes:
Attribute | Description |
|---|---|
ID: | The "ID" attribute represents the unique identifier of the send task. It corresponds to the task definition key in the BPMN model. |
Name: | The "Name" attribute defines the name or label of the send task. |
Implementation
The "Implementation" attribute specifies the method, class, or service that the Send task will invoke or execute. It defines the technical implementation behind the task's functionality. Implementation options include:
Implementation | Description |
|---|---|
Expression: | An expression or script can be used to define the task's behavior. This allows for flexible and custom task execution based on the defined expression. Result Variable: You can specify the name of the variable where the task result will be stored. |
Asynchronous Continuations
Use asynchronous continuations to control whether a task runs asynchronously:
Attribute | Description |
|---|---|
Asynchronous Before: | The "Asynchronous Before" attribute indicates if the task can be started asynchronously before other tasks or events in the process. |
Asynchronous After: | Similarly, the "Asynchronous After" attribute denotes whether the send task can be completed asynchronously after other tasks or events in the process. |
Service Tasks Configuration
A Service Task represents work performed by a service or external system as part of a business process. Service tasks are typically used to model automated activities that interact with external services, applications, or systems rather than requiring human interaction.
Overview
A service task is a specialized workflow activity that executes custom logic written in JavaScript or TypeScript. Service tasks run automatically on the server side without any user intervention.
In a BPMN workflow, various actors may participate in the process, including APIs or systems that require decision-making capabilities. Service tasks handle these automated interactions.
Purpose
The primary purpose of a service task is to automate backend operations within an agentic workflow. It acts as the execution engine of your workflow, handling data manipulation, system integrations, and computational logic.
Key capabilities include:
System Integration: Calling external APIs (REST) to fetch or push data.
Data Transformation: Processing process variables, formatting dates, or calculating values.
Configure and Code
Double-click on the service task to select it.
In the pop-up panel, click Create/Choose Service. This opens the Service Editor.
Name your Service: Give it a unique identifier (e.g.,
calc_refund_svc).Select Language: Choose TypeScript or JavaScript.
Click Save. A code editor opens:
Write Logic: Enter your custom code in the main function. This function receives three key arguments:
task: The current task instance; use this to get variables viatask.variables.get("name").taskService: Methods to complete or fail the task.processVariables: Methods to set or update the workflow's data payload.
Click SAVE.
User Tasks Configuration
A User Task in a BPMN journey is an activity that requires human intervention in a business process. When a process instance reaches a user task, a job is created, and the process instance pauses until the job is completed. Users can be automatically assigned to these tasks, or a job worker can manually subscribe to them.
Attributes
User task attributes:
Attribute | Description |
|---|---|
ID: | The "ID" attribute is the user task's unique identifier. It corresponds to the task definition key in the BPMN model. |
Name: | The "Name" attribute defines the name or label of the user task. |
Assignments
Assignments determine who can handle the task and can include the following attributes:
Attribute | Description |
|---|---|
Assignee: Default: | The "Assignee" specifies the user who is assigned to the task. When values need to be determined dynamically during task execution, they're represented as Expression Language (EL) expressions. By default, it is |
Candidate Users: | The "Candidate Users" attribute can be used to specify a list of users who are potential candidates for completing the task. This field may remain empty or be populated with user identifiers based on the process requirements. |
Candidate Groups: | The "Candidate Groups" attribute allows you to specify groups of users who are eligible candidates for the task. This field may contain one or more group identifiers based on the process design. |
Scheduling
User tasks support specifying a task schedule, which helps define when users should interact with the task and includes:
Attribute | Description |
|---|---|
Follow Up Date: | The "Follow Up Date" attribute specifies the date when a follow-up action should be taken for the task. It can be expressed as an EL expression (e.g., |
Priority: | The "Priority" attribute can be used to set the task's priority level. The specific values for priority levels may be defined in accordance with the business process requirements. |
Due Date: | The "Due Date" attribute specifies the deadline for completing the user task. Similar to the "Follow Up Date," it can be expressed as an EL expression (e.g., |
Asynchronous Continuations
Use asynchronous continuations to control whether a task runs asynchronously:
Attribute | Description |
|---|---|
Asynchronous Before: | The "Asynchronous Before" attribute indicates if the task can be started asynchronously before other tasks or events in the process. |
Asynchronous After: | Similarly, the "Asynchronous After" attribute denotes whether the user task can be completed asynchronously after other tasks or events in the process. |
Webhook Tasks Configuration
A Webhook Task represents workflow activities that interact with external systems requiring asynchronous callbacks before the workflow can continue. Webhook tasks are typically used to model activities that initiate an outbound request to an external platform and then pause execution until a callback response is received, rather than requiring human interaction.
Webhook task attributes:
Attribute | Description |
|---|---|
Name | The "Name" attribute defines the name or label of the webhook task. |
Request | The "Request" attribute specifies the server-side function responsible for initiating the outbound request to the external system. This function triggers the external event and begins the asynchronous operation. |
Handler | The "Handler" attribute specifies the server-side function responsible for processing the callback response received from the external system once the external event is completed. Once invoked, the handler captures the result and updates the workflow state so that execution can continue. |
Managing Workflows
All existing workflows are accessible from the Agentic Workflows page, where you can view, edit, deploy, and execute them as needed.
Editing a Workflow
To improve workflow efficiency, the platform lets you update journeys directly from the Agentic Workflows page. This quick-access editing capability lets you make configuration changes on the spot, streamlining your workflow development process.
To update a journey:
Go to the Home page > Agents > Agentic Workflow.
The Agentic Workflows page opens and displays all workflow journeys in card view.
Enter the workflow journey name in the Search box to filter it.
Click the Ellipsis
icon on the workflow journey card.Click Edit.
The Workflow Canvas page opens, allowing you to view and modify your workflow. Refer to Creating a Workflow Diagram for detailed instructions.
If the canvas is in read-only mode, you must manually activate edit mode to proceed with modifications.
Click CHECKOUT on the top-right corner of the canvas to enable editing.
The CHECKOUT confirmation box opens.
Click CHECKOUT.
The platform switches the canvas from read-only mode to edit mode, allowing you to make changes. Refer to Creating a Workflow Diagram for detailed instructions.
Deploying a Workflow
Once you have created and finalized your workflow journey, the next step is deploying it to the runtime platform. The runtime platform is a separate environment that lets you test and execute your workflows, ensuring all processes run as designed.
There are two deployment modes:
Live - Use this for testing your workflow in a development environment. Any changes you make to workflow task properties are automatically applied to the runtime, allowing you to iterate and refine your workflow seamlessly.
Production - Use this to execute your workflow in a production environment for real-world operations.
Caution
Each app can have only one workflow journey deployed at a time.
To deploy a workflow:
Go to the Home page > Agents > Agentic Workflow.
The Agentic Workflows page opens and displays all workflow journeys in a card view.
Enter the workflow journey name in the Search box to filter it.
Click the Ellipsis
icon on the workflow journey card.
Click Deploy.
The Deploy Workflow confirmation window opens.
Choose your deployment mode from the dropdown:
Live - For testing the execution
Prod - For the operational execution
Click Deploy.
A success message appears confirming deployment.
Now that you have deployed your workflow, you can use Agent Evaluation to evaluate its performance against curated test scenarios and measure real outcomes, such as task completion rates, output quality, correctness of tool usage, and performance.
Running a Workflow
Running a workflow is the critical process of executing your designed workflow in a runtime environment separate from the design interface.
The runtime testing process serves multiple important purposes:
You can observe how your workflow executes in practice, verifying that tasks are executed in the correct sequence, data flows properly between tasks, and conditional logic functions as intended.
You can identify any issues, errors, or bottlenecks that may not be apparent during the design phase but become evident during actual execution.
You can analyze results to ensure the workflow produces the expected outcomes.
Based on the runtime process, execution results, and any challenges encountered, you can refine and improve your workflow to enhance its effectiveness. This iterative testing and refinement process ensures that your workflow functions correctly and efficiently before deploying it to production, where it will handle actual business operations.
To run a workflow:
Go to the Home page > Agents > Agentic Workflow.
The Agentic Workflows page opens and displays all workflow journeys in a card view.
Enter the workflow journey name in the Search box to filter it.
Click Run on the workflow journey card.
The runtime environment opens automatically in a new window.
Click the Add + icon at the bottom-right corner of the page.
A list of all deployed workflows appears in the right-side panel.
Enter the workflow name in the Search box to filter results.
Select the workflow you want to execute.
Click Proceed to initiate execution.
Once started, you can track and view the status of each task execution in real-time as the workflow progresses.
Suspending, Resuming, and Terminating Instances
You can suspend, resume, and terminate instances via the buttons at the top of the main panel.
To suspend an instance, click the Suspend button, then click Suspend on the confirmation dialog:
To resume an instance, click the Resume button, then click Resume on the confirmation dialog:
To terminate an instance, click the Terminate button, then click Terminate on the confirmation dialog:
Caution
This action is irreversible. A terminated instance cannot be recovered or resumed.
Terminating an instance permanently deletes it from the workflow engine and marks it as Externally Terminated. Use terminate only when you are certain the instance should no longer run. Terminated instances are moved to the Completed tab for audit visibility.
Viewing a Workflow Status History
Monitor the status of your workflow executions in real time by viewing comprehensive details for each run, including the workflow name, current execution status, timestamp, and a flowchart visualization.
To access the workflow's status history:
Click the Run History
icon on the left side menu.The History page opens and displays the latest details of the workflow execution status.
The left side panel organizes all workflow runs into two tabs:
Ongoing - Displays all suspended and currently executing workflows. Suspended workflows have an orange suspended
icon.Completed - Displays all workflows that have finished execution or that have been externally terminated. Completed workflows have a green checkmark
icon, and terminated workflows have a red canceled 
icon.
Use the search box and filters to quickly locate a specific workflow run, and then select to view the details in the right panel across these tabs:
Timeline View tab - Select to view execution progress with timestamped tasks and color-coded statuses.
Detailed View tab - Select to view the complete workflow diagram with its current execution status.
Manage Variables tab - Select to view all data and variables that were created during the workflow execution process. You can manage these variables as needed for debugging or analysis.
Incident tab - When an instance is suspended due to an external task failure, the Incident Detail tab appears in the right panel.
The Incident Detail tab shows:
Field
Description
Incident Type
Category of the failure (e.g.
failedExternalTask)Incident Message
Human-readable error description
Configuration ID
ID of the failed external task
Incident Time
Timestamp when the incident was recorded
Viewing the Workflow Execution History
The platform enables you to monitor and analyze agentic workflow performance through comprehensive execution information. Detailed execution logs show how your workflows process requests and generate results, helping you make informed optimization decisions.
Note
Execution history tracks only Service tasks, Action tasks, and AI Agent tasks.
To view the workflow execution details:
Click the Ellipsis
icon on the workflow journey card.
Click View Last Run.
A right panel opens, displaying metadata about the last execution, including input/output, execution sequence, time, and other performance metrics.
Click View Run History.
A right-side panel opens, displaying all its execution records for the last five runs.
To view the previous or next execution record, click << or >>.
