Documentation
Start typing to search…
  1. Integrations
  2. Technology

Browser Run

Overview

Browser Run is a native capability within Leena AI's Workflows Studio that enables automated web-based actions through a headless browser running on a virtual machine. This powerful feature allows organizations to automate complex web interactions that would typically require manual intervention by administrators or IT personnel.

Key Capabilities

  • Headless Browser Automation: Executes web actions without requiring a visible browser interface
  • Admin-Level Access: Performs actions using administrative credentials on behalf of employees
  • Cross-Platform Compatibility: Works with any web-based application or system
  • Multiple Parallel Task Executions: Executes multiple tasks in parallel

How It Works

Let's walk through a complete example of applying for leave in SAP SuccessFactors to understand how Browser Run operates:

Task Input

Apply for a leave request on behalf of user John Smith:

Step 1: Task Interpretation and Planning

The AI agent receives the natural language instruction and:

  • Parses the objective: Apply vacation leave for John Smith
  • Identifies required data: Leave type, dates, credentials, target system.
  • Creates execution plan: Login → Navigate → Fill form → Submit → Verify
  • Validates inputs: Ensures date format, leave type validity

Step 2: Browser Initialization and Navigation

  • Launches headless browser on secure VM with isolated session
  • Navigates to SAP URL: https://company.successfactors.com
  • Takes initial screenshot for audit trail
  • Waits for page load and dynamic content rendering

Step 3: Authentication Process

  • Computer vision scans page to identify login elements
  • Locates username field (even if ID/class names change)
  • Enters credentials: [email protected] in username field
  • Finds password field and enters AdminPass123
  • Identifies and clicks "Login" or "Sign In" or "Continue" button
  • Waits for authentication and verifies successful login

Step 4: Homepage Navigation and Menu Discovery

  • Analyses loaded homepage using computer vision
  • Scans for navigation elements: menus, buttons, links
  • Identifies leave-related options: "Request Time Off", "My Leave", "Employee Services"
  • Clicks on "Request Time Off" button (adapts to button text variations)
  • Handles any intermediate pages or redirections automatically

Step 5: Form Interaction and Data Entry

  • Analyses leave request form structure and fields
  • Locates leave type dropdown:
  • Computer vision identifies dropdown element
  • Clicks to expand options
  • Searches for "PTO" or similar terms
  • Selects appropriate option
  • Handles start date field:
  • Identifies date picker or input field
  • Enters 2024-12-22 in correct format
  • Handles calendar popup if present
  • Handles end date field:
  • Enters 2024-12-23 using the same logic
  • Validates date range logic

Step 6: Form Submission and Validation

  • Reviews form completion: Ensures all required fields filled
  • Locates submit button: "Submit", "Apply", "Send Request"
  • Clicks submit button with human-like timing
  • Waits for processing: Handles loading screens and delays
  • Monitors for validation errors: Checks for error messages or field highlighting

Step 7: Confirmation and Result Capture

  • Identifies success indicators:
  • Success message: "Leave request submitted successfully"
  • Reference number: "Request ID: LR-2024-001234"
  • Confirmation page elements
  • Takes confirmation screenshot for audit purposes
  • Extracts key information: Request ID, status, next steps
  • Prepares result summary with execution details

Step 8: Result Reporting

Browser Run returns structured results:
json

"status": "SUCCESS", "requestId": "LR-2024-001234", "message": "Leave request submitted successfully", "details": "employee": "John Smith", "leaveType": "PTO", "startDate": "2025-12-22", "endDate": "2025-12-23", "duration": "2 days" , "screenshots": ["login.png", "form.png", "confirmation.png"], "executionTime": "2m 34s"

Setup Connector

There is no requirement of setting up a connector for 'Browser Run' as this is a native Leena AI's Workflow Studio capability.

Configuration

This guide provides step-by-step instructions for setting up and configuring Browser Run functionality within Leena AI's Workflows Studio. 'Browser Run' allows you to automate web-based tasks through a headless browser, such as applying for leave requests in SAP SuccessFactors on behalf of users.

Prerequisites

Before configuring Browser Run, ensure you have:

  • Access to Leena AI Workflows Studio
  • Valid credentials for the target web application
  • Understanding of the target web application's interface and workflow

Step-by-Step Setup Process

Step 1: Create a New Workflow

  1. Navigate to Workflows Studio in your Leena AI platform
  2. Go to Apps and create or select your application
  3. Click on the Workflows tab
  4. Create a new workflow or edit an existing one

Step 2: Add an Action Node

  1. In the workflow canvas, locate the Add nodes panel on the left side
  2. Click on the Action node type from the available options
  3. Drag and drop the Action node into your workflow at the desired position
  4. Connect the Action node to your workflow sequence using the connection points

Step 3: Configure Action Node Properties

  1. Click on the Action node to open the Action Node Properties panel
  2. Fill in the basic node information:
    • Node Title: Enter a descriptive title (e.g., "Apply Leave via Browser Run")
    • Node Description: Provide a detailed description of what the action will do
    • Node ID: Auto-generated from node title

Step 4: Connect to Browser Run App

  1. In the Connect App section:
    • Select App: Choose "Browser Run by Leena" from the dropdown menu
    • Select Connection: This is auto-populated (no user configuration required)

Step 5: Select Browser Action

  1. In the Select Action section:

    • Choose your specific browser action (e.g., "Apply Leave in SAP SuccessFactors")

Step 6: Define Task Instructions

In the Task field, provide detailed instructions for the browser automation. Structure your instructions clearly:

Example Task Configuration

**Objective:** Apply for a leave request on behalf of the user using the SAP SuccessFactors portal.

**Leave Details:**
* Leave Type: {SAP_LEAVE_TYPE}
* Start Date: {SAP_LEAVE_START_DATE}
* End Date: {SAP_LEAVE_END_DATE}

**Login Credentials:**
* URL: {SAP_URL}
* Company ID: {SAP_COMPANY_ID}
* Username: {SAP_USERNAME}
* Password: {SAP_PASSWORD}

**Detailed Steps:**
step 1: Navigate to the specified URL: {SAP_URL}
step 2: Log in using the provided credentials
step 3: Once logged in, if the homepage/dashboard is not displayed, re-open the same URL in a new browser tab
step 4: Locate and click on the "Request Time Off" button
step 5: In the time-off request interface, select the appropriate Time Type, Start Date, and End Date
step 6: Click Submit to finalize the leave request
step 7: Wait for the success message and capture the confirmation

Step 7: Configure Metadata and AI Settings

  1. Add Metadata (Optional):

    • Click "Add Metadata" to enter key-value pairs for all the keywords provided under 'Task' such as, SAP_COMPANY_ID
    • Note: This metadata will not be sent to the LLM for security purposes

  2. Callback URL:

    • Enter the callback URL to receive task results

  3. LLM Provider:

    • Select the AI model for task execution (e.g., "Gemini")
    • This AI model will interpret and execute the browser automation instructions

Step 8: Save and Test

  1. Click Save as draft to save your configuration without activating
  2. Alternatively, click Save to save and activate the workflow
  3. Test the workflow with sample data to ensure proper functionality

Updated about 19 hours ago