Industry: Corporate Services
Region: USA
Technology: Generative AI, Natural language processing (nlp), document processing
About the Client
Our client, a leading corporate services provider, involved in generating invoices for customers and managing payments made to Bank.
Business Challenges
Out Client, provider of invoice generation and processing services, faced significant challenges in matching payment data received from Bank with the corresponding invoice details sent via Outlook emails. The primary issues included.
- Inconsistency in Entity Names: Names on Bank's portal often, not matching exactly with the detailed entity names received via email.
- Manual Matching Process:Previously, matching entities was a manual and error-prone task.
- Inefficiency:leading to inefficiencies and potential inaccuracies in payment reconciliation and impacting overall payment processing.
The Solution
Our Evoke experts designed a comprehensive solution:
- AI Automated Invoice Detail Extraction:Our solution leverages the documentsolutions.ai to automate the extraction of crucial invoice-related entities such as the customer's entity name, invoice amount, and invoice number from Outlook emails. These details are then stored in a SQL database, ready for further processing.
- Order based Entity Mapper: This layer filters the data based on the payment amount and then performs n-grams matching, considering the sequence of words and cross-verifying against sender email domains for additional context.
- Max Matching Entity Ranker:After the previous layer, if there is still ambiguity in the resulting entities, then this layer is performed to score potential matches by applying the custom ranking method which prioritizes the n-gram matchings by deprioritizing their order.
- Semantic Entity DetectorIf ambiguity persists still, this layer uses a documentsolutions.ai to semantically analyze and compare potential matches, resolving any ambiguities that might have arisen in earlier stages and ensuring the best possible match.
The Benefits
The implementation of this AI system has transformed our client’s payment reconciliation processes, delivering significant benefits:
- Increased Efficiency:The automation of entity matching has significantly reduced the time previously spent on manual processes, allowing employees to focus on higher-value activities.
- Enhanced Accuracy:The use of advanced AI techniques reduces the potential for human error and ensures highly accurate matching of payment and invoice data.
- Improved Customer Satisfaction:Faster processing and more accurate invoice processing System contribute to better overall service, enhancing customer satisfaction and trust.
Key Technologies
- Python
- Azure OpenAI