Edit

Automate bill copy requests with Power Platform

Many organizations still rely on manual processes to retrieve and send bill copies, especially when working with legacy systems that lack modern integration options. This reliance often requires account representatives to navigate multiple screens, search for invoices, and respond to requests individually, leading to delays and inefficiencies.

This article shows how Microsoft Power Platform can automate bill copy requests. This solution replaces manual effort with a scalable, reliable solution that improves processing speed and provides better visibility into operations.

Tip

This article provides an example scenario and visual representation of how to automate bill copy requests using Power Platform. This solution is a generalized example scenario architecture, which can be used for many different scenarios and industries.

Architecture diagram

Architecture diagram showing how bill copy requests flow from a Power Apps model-driven app into Microsoft Dataverse and are processed using Power Automate (cloud and desktop flows), with results visualized in Power BI.

Workflow

Microsoft Power Platform provides the foundation for building a secure, scalable, and automated solution for managing bill copy requests through apps, automation, and reporting.

  1. Create bill copy requests in Power Apps:

    • Customer service representatives create bill copy requests in a Power Apps model-driven app.
    • Bill copy requests are standardized to improve data integrity and reduce manual tracking errors.
    • Dataverse is used as the central repository for all bill copy request data.

  2. Process bill copy requests with Power Automate:

    • A scheduled Power Automate cloud flow triggers Power Automate desktop flows at defined intervals.
    • The desktop flow accesses bill copy requests in Dataverse, searches for invoices in bank-provided web-based applications, and processes the requests automatically.
    • Customer service representatives receive real-time notifications confirming successful processing.

  3. Monitor bill copy request metrics with Power BI:

    Built on top of Dataverse, a Power BI dashboard provides leadership with visibility into:

    • Daily, weekly, and monthly request volumes
    • Processing success rates
    • Trends in customer demand

    This visibility enables data-driven decision-making and improved resource allocation.

Components

Scenario details

This architecture demonstrates how Microsoft Power Platform can modernize operations in the energy and utility sector, reduce costs, and improve customer satisfaction.

Business challenge

An energy and utility organization receives more than 100 daily customer requests for bill copies, creating an operational bottleneck.

The organization used a bank-provided web-based application to search for, locate, and send bill copies. However, the process was manual and tedious. Account representatives had to navigate each account, filter by date, identify invoices, and send them individually. Each request took 10 to 30 minutes, depending on the number of invoices involved. This process created inefficiencies, delayed responses, and affected customer satisfaction.

Business value

  • After the organization built and deployed the solution, each request completed in about 30 seconds instead of 10 to 30 minutes.

  • By combining model-driven apps, Power Automate cloud and desktop flows, and Power BI, the organization transformed a labor-intensive process into a fully automated workflow.

  • With AI-driven enhancements on the horizon, the solution is poised to deliver even greater value.

Prerequisites

To enable this solution, use the following licenses:

  • Power Automate premium license (per user):

    • Assign to account representatives so they can securely access and use the model-driven app.
    • Ensure they can log requests into Dataverse.

  • Power Automate premium license (service account):

    • Use a dedicated service account to run scheduled flows.
    • Provide secure, centralized automation without tying execution to individual users.

  • Power Automate unattended robotic process automation (RPA) license:

    • Enable Power Automate desktop to run on a virtual machine in unattended mode.
    • Allow the automation to execute without human intervention, ensuring requests are processed reliably and consistently.

This licensing model balances cost efficiency with scalability, ensuring compliance while maximizing automation benefits.

Technical constraints and trade‑offs

  1. Legacy system limitations

    Constraint:

    • The legacy web application doesn't expose REST or SOAP APIs, webhooks, or connector support.
    • All required operations, like reading data and sending emails, are only available through the UI.

    Trade-off:

    • RPA is considered the only viable integration method.
    • UI automation introduces sensitivity to layout changes and system latency, so you need robust retry logic and checkpointing.

  2. Single‑threaded bot execution

    Constraint:

    • Power Automate desktop bots execute one session at a time per machine.
    • No native parallelism within a single bot session.

    Trade-off:

    • Throughput is limited by sequential execution.
    • However, because the daily volume (100 requests or fewer) is well below the bot's capacity (about 200 per day), this constraint doesn't affect service level agreements (SLAs).
    • Machine groups were available but unnecessary, avoiding extra operational overhead.

  3. Credential security requirements

    Constraint:

    • You can't store credentials locally, embed them in flows, or expose them to developers.
    • Enterprise security requires centralized secret management.

    Trade-off:

    • Use Azure Key Vault and Dataverse credential objects.

    • This approach adds initial configuration complexity but ensures:

      • Zero credential exposure
      • Automated rotation
      • Least‑privilege access
      • Full auditability

  4. Service account permission boundaries

    Constraint:

    • Restrict the dedicated service account to:
      • Read-only access in the legacy system
      • Send email permissions
      • No delete, write, or administrative privileges

    Trade-off:

    • Ensured data integrity and minimized risk.
    • Required the automation to operate strictly within the allowed UI paths.
    • Prevented certain optimizations that would require elevated permissions.

  5. UI-driven automation sensitivity

    Constraint:

    • UI automation is inherently sensitive to:
      • Page load delays
      • Element rendering problems
      • Browser updates
      • Unexpected pop-ups

    Trade-off:

    • Implemented robust retry logic and Dataverse-based checkpointing.
    • Added overhead to flow design but improved resilience.
    • Ensured failed or partial runs could resume without manual intervention.

  6. Dataverse as the state and logging layer

    Constraint:

    • All request statuses, checkpoints, and error logs had to be persisted in Dataverse.

    Trade-off:

    • Introduced extra read and write operations per transaction.

    • Provided a reliable, centralized state store that enables:

      • Automatic reprocessing of failed or partial requests
      • Idempotent execution
      • Full audit trail

  7. Infrastructure and maintenance windows

    Constraint:

    • Bot machines require periodic operating system (OS) updates, browser updates, and Power Automate Desktop (PAD) version updates.

    Trade-off:

    • Slightly reduces available runtime.

    • Mitigated by:

      • Scheduling updates during off‑peak hours
      • Maintaining throughput headroom (50% or less utilization)

Considerations

These considerations implement the pillars of Power Platform Well-Architected, a set of guiding tenets that improve the quality of a workload. Learn more in Microsoft Power Platform Well-Architected.

Reliability

This solution includes multiple reliability safeguards to ensure that unattended Power Automate desktop executions stay stable and recoverable. These safeguards make the process resilient to UI variability, system latency, and transient failures. They work across credential security, execution monitoring, state tracking, and automated recovery.

Scheduled maintenance and update controls

Machines follow a controlled update cycle.

Safeguards:

  • Apply OS patches and PAD updates during off-peak windows
  • Validate browser updates before rollout
  • Monitor for breaking UI changes

Outcome: Reduces unexpected downtime and maintains long-term reliability.

Centralized logging and observability

The solution captures all execution details for audit and troubleshooting.

Safeguards:

  • Dataverse logs for request-level tracking
  • PAD logs for step-level execution
  • Key Vault logs for credential access
  • Cloud flow logs for orchestration events

Outcome: Full traceability and rapid root-cause analysis.

Operational Excellence

Time savings: Reduced processing time per request from up to 30 minutes to just 30 seconds.

Efficiency gains: Automated handling of more than 100 daily requests without manual intervention.

Transparency: Leadership can monitor performance and trends in real time.

Employee productivity: Account reps are freed from repetitive tasks, so they can focus on higher-value customer interactions.

Next steps

You can extend this solution with AI integration using AI Builder.

Architecture diagram showing bill copy requests from Outlook processed with AI Builder and Power Automate, stored in Microsoft Dataverse, automated with RPA, and visualized in Power BI.

  • The solution reads bill copy requests directly from a shared mailbox.
  • AI Builder automatically extracts key details, such as the account number, date range, and invoice type.
  • The solution logs requests into Dataverse without any human input.

This enhancement eliminates manual request entry to improve efficiency and reduce costs.

Contributors

Microsoft maintains this article. The following contributors wrote this article.

Principal authors:

  • Last updated on