Skip to Content
Agent Workflows

Agent Workflows

This comprehensive guide explains how auditors can set up workflow automations that run automatically when clients upload documents to requests. The system provides powerful AI-driven capabilities for document processing, data extraction, and validation.

Workflow automation transforms manual document processing into an efficient, accurate, and scalable operation that runs automatically whenever clients submit documents.

Overview

Workflow automations allow you to:

  • Automatically process client uploads with AI-powered document analysis
  • Extract structured data from unstructured documents
  • Match evidence across different data sources
  • Validate client submissions against expected formats
  • Perform deep analysis with custom business rules

⚙️ Workflow Onboarding

Initial configuration and data structure setup to define how your automation will work.

🎯 Attribute Configuration

Detailed automation rules and processing parameters for fine-tuning behavior.

Getting Started: Workflow Onboarding

Initial Decision Point

When setting up a new workflow, you’ll first choose between two primary approaches:

Best for: Requests involving itemized data, samples, detailed schedules

Use Cases:

  • Bank reconciliations
  • Inventory listings
  • Transaction details
  • Account schedules

What happens: Creates a structured table that client uploads will be matched against

Perfect for financial data where you need to match uploaded information against expected formats and perform calculations or comparisons.

Workflow Capabilities

The system automatically enables different capabilities based on your task configuration:

🔧 Core Automation Features

  • • Match Evidence: Automatically link uploaded documents to expected data points
  • • Extract Data with AI: Pull structured information from unstructured documents
  • • Perform Deep Analysis: Apply business rules and validation logic

🎯 Client Request Features

When your task is visible to clients, you also get:

  • • Validate Client Uploads: Automatically check submissions against requirements

Detailed Setup Process

Create Sample Table

Click “Add Table” to begin structured data setup

Input Sample Data:

  • Copy data from Excel/CSV
  • Use the built-in table editor
  • Format columns and headers appropriately

Review Structure: Ensure column headers match expected client data

Document Group Naming

Provide descriptive names using clear, specific terminology

Examples:

  • “Bank Statements Q3 2024”
  • “Inventory Count Sheets”
  • “AP Aging Detail”

System Features:

  • Name validation for valid characters and length
  • Auto-capitalization of first letters

Matching Configuration

Configure how the automation will process uploads

Setup Tasks:

  • Source Selection: Choose existing data sources to match against
  • Field Mapping: Configure how uploaded data maps to your sample structure
  • Matching Rules: Set up fuzzy matching, variance tolerance, required fields
  • Review and Test: Validate matching logic before activation

Advanced Configuration: Add Edit Data Source

After completing workflow onboarding, you enter the detailed configuration phase where you define specific automation behaviors.

Advanced Configuration: This phase requires careful attention to detail as it determines how accurately your automation will process client uploads.

Core Processing Configuration

Define how data is extracted and validated

  • Define Extraction Fields: Specify what data to extract from documents
  • Set Data Types: Configure field types (text, number, date, currency)
  • Validation Rules: Establish acceptance criteria and validation logic
  • Required vs Optional: Mark critical fields that must be present

Best Practice: Start with the most critical fields first, then add optional fields as your workflow matures.

Automation Execution Flow

📋 Upload Processing Pipeline

Document Receipt

Client uploads files to request

OCR Processing

Documents scanned and text extracted

Structure Analysis

AI analyzes document layout and content

Data Extraction

Relevant information pulled based on configuration

Validation Checks

Data verified against established rules

Matching Logic

Information matched to existing data sources

Results Generation

Processed data made available for review

Progress Monitoring Features

  • Real-time progress updates during processing
  • Detailed status messages for each processing stage
  • Error reporting and exception handling
  • Completion notifications and summary reports

Configuration Best Practices

For data-driven automation setups

1. Sample Quality

Create comprehensive, representative sample data that covers all expected variations

2. Column Headers

Use clear, consistent naming conventions that match client terminology

3. Data Types

Ensure sample data reflects expected client format (dates, currencies, numbers)

4. Coverage

Include all variations you expect to receive from different clients

Common Workflow Scenarios

🏦 Bank Reconciliation

Setup: Structured Data (Add Table)

  • • Sample: Date, Description, Debit, Credit, Balance
  • • Match against GL export data
  • • Set variance tolerance
  • • Auto-flag unmatched items

📄 Invoice Processing

Setup: Document-Only (No Table)

  • • Create “Vendor Invoices” group
  • • Extract: Vendor, Date, Amount, PO
  • • Set validation rules
  • • Match against vendor lists

📋 Compliance Review

Setup: Document-Only (No Table)

  • • Category: “ISO Certifications”
  • • Extract metadata (dates, scope)
  • • Monitor expiration dates
  • • Track compliance status

Troubleshooting Common Issues

Common configuration and setup issues

❌ “No datasources were affected by this query”

Cause: Configuration doesn’t match any processing rules

Solution: Review attribute configuration and ensure proper field mapping

⏱️ Processing Timeouts

Cause: Large file volumes or complex processing rules

Solution: Break into smaller batches or simplify processing logic

🎯 Poor Matching Results

Cause: Inadequate sample data or overly strict matching rules

Solution: Expand sample data and adjust tolerance settings

Monitoring and Maintenance

📊 Performance Tracking

⏱️

Processing Times

✓

Success Rates

⚠️

Error Patterns

😊

Client Satisfaction

Continuous Improvement

  • Rule Refinement: Regularly update matching and validation rules
  • Sample Updates: Expand sample data based on actual client submissions
  • Process Optimization: Streamline workflows based on usage patterns
  • Technology Updates: Leverage new AI capabilities as they become available

Success Metrics

Measuring automation performance

📈 Processing Accuracy

Percentage of correctly processed uploads

⏰ Time Savings

Reduction in manual processing time

🎯 Error Reduction

Decrease in processing mistakes

⚡ Response Time

Faster feedback on submissions

Integration with Audit Workflow

🔗 Task Management Integration

  • • Status Updates: Automatic progress reporting in task interface
  • • Exception Routing: Failed processes routed to appropriate team members
  • • Documentation: Complete audit trail of all processing activities
  • • Reporting: Automated generation of processing summaries

💬 Client Communication

  • • Automated Responses: Immediate acknowledgment of successful uploads
  • • Exception Notifications: Clear communication when issues arise
  • • Progress Updates: Real-time status information for clients
  • • Completion Confirmations: Automated notification when processing complete

The workflow automation system transforms manual document processing into an efficient, accurate, and scalable operation. By following this guide and implementing appropriate configurations, audit teams can significantly improve their efficiency while maintaining high quality standards and providing excellent client service.

Advanced Tips

Building reusable workflow configurations

  • Save Successful Configurations: Create templates for common workflow types
  • Standardization: Use consistent naming and configuration patterns
  • Documentation: Maintain clear records of configuration decisions
  • Knowledge Sharing: Share effective configurations across audit teams

Template Strategy: Start with one successful workflow and create variations for different client types or engagement sizes.

Last updated on
System Settings