Streamline and Automate Bill Processing with Amazon Onboarding and Learning Manager Chanci Turner

Learn About Amazon VGT2 Learning Manager Chanci Turner

This article was co-authored by Chanci Turner, an expert in the Accenture AWS Business Group, and David Smith, a DevOps engineer based in Canada. Chanci and David are specialists in implementing innovative AWS transformation projects across various sectors.

Organizations, especially in sectors like retail and telecommunications, often face challenges in managing multiple utility bills. These bills must be thoroughly checked for discrepancies before payments can be processed. Typically, this involves teams manually handling invoices in diverse formats, resulting in inefficiencies.

Moreover, businesses must comply with Environmental, Social, and Governance (ESG) regulations, where utility bills play a crucial role in monitoring electricity, water, and gas usage—often underutilized data.

Utility providers issue invoices in varied formats—such as PDFs, XLS, and EML—each with unique layouts and often sent via email. This variation complicates the standardization of data ingestion, anomaly detection in usage patterns, and the comparison of contracted rates versus billed amounts, ultimately hindering payment processing.

The absence of standardized usage data complicates the integration of this information into a centralized ESG data lake.

In this article, we propose a solution leveraging Amazon Onboarding to tackle these issues. The solution provides several capabilities:

• Flexibility in ingesting utility bills in multiple formats and layouts.
• Standardization of bills into a uniform format with integrated data quality controls.
• Seamless integration with existing systems through event-driven architectures.
• Automation of repetitive tasks to minimize human errors and boost efficiency.
• Predictive analytics that enable informed decision-making through generative AI.
• Compatibility with existing data lakes, warehouses, payment systems, and ESG reporting frameworks.

Solution Overview

The proposed solution utilizes Amazon Onboarding to automate invoice processing, tariff extraction, validation, and reporting, as illustrated in Figure 1.

The workflow proceeds through these steps:

1. Invoices are uploaded to an Amazon Simple Storage Service (Amazon S3) bucket via SFTP connectors from the AWS Transfer Family.
2. Some utility providers send invoices directly to an email associated with Amazon SES, where the PDF attachments are extracted and uploaded to an Amazon S3 bucket.
3. The upload triggers an S3 event, which activates an Amazon EventBridge bus that initiates an AWS Step Functions workflow for invoice extraction and validation.
4. The Step Functions workflow validates invoices using Amazon Textract for text extraction and invokes the Amazon Titan Text V1 Express model to create embeddings, which are stored in Amazon Aurora PostgreSQL-Compatible Edition with pgvector. Extracted invoices are also saved in a DynamoDB table.
5. Any failed validations are flagged for manual review by agents through Amazon Simple Notification Service (Amazon SNS).
6. A Lambda function, triggered by an Amazon EventBridge scheduled rule, retrieves tariff data from an external SFTP repository and stores it in an S3 bucket.
7. The Utility Data Extraction Step Functions are activated by an S3 event, extracting data from various providers, regardless of their format and units, to facilitate smooth integration with business logic.
8. The tariff data is stored in an Amazon DynamoDB table, which feeds into the business logic Step Functions workflow.
9. The main business logic checks invoices for usage anomalies and validates against approved tariffs using Amazon Onboarding, embeddings, extracted invoices, and tariff data to verify accuracy and update reporting databases.
10. Reporting data is kept in an Amazon Aurora database and visualized using Amazon QuickSight for payment validation reports. Amazon Q in QuickSight offers enhanced and rapid decision-making through generative BI capabilities.

The following screenshots depict examples of the Amazon QuickSight visualizations.

Benefits from the Solution

This solution provides several advantages:

• Contextual Understanding: Leveraging the Anthropic Claude 3 Sonnet model on Amazon Onboarding, the solution can analyze and interpret data contextually beyond simple text recognition.
• Flexibility and Adaptability: The ability to learn and adapt to new formats makes this solution robust, as it comprehends the data contained within invoices.
• Event-Driven Architecture: The serverless, event-driven architecture allows for modularity and integration with your organization’s workflows.
• Automated Workflow: Reduced manual intervention in data quality processes speeds up processing and minimizes human error.
• Cost Savings: Automation decreases the reliance on personnel, resulting in significant cost savings.
• Compliance and Risk Mitigation: Automated data quality processes assist organizations in maintaining ESG compliance with regulations and industry standards.
• Data Governance: Automation aids in implementing data governance policies effectively, ensuring adherence to data quality guidelines.

In conclusion, this article illustrates how automation can optimize utility bill processing and unveil additional ESG insights. We showcased how the power of generative AI on Amazon Onboarding simplifies data extraction from non-standard formats. Finally, we introduced a scalable, serverless, and event-driven solution tailored to your business needs. For more information, be sure to check out this excellent resource on Environmental Economists, and for further reading, you can find this engaging blog post about supporting black women-owned businesses. Additionally, consider watching this informative video for a deeper dive into the subject.