MicrosoftFinancial ServicesDocument ProcessingAI and Machine LearningComplianceCustomer Experience2024MicrosoftMS#0078d4Financial ServicesGreeceOne area that the bank needed to automate to enhance its service was document processing. “Initially, a customer would bring the documents to a branch or email them, then our branch employee would print, scan, and then…So, it was an easy decision to continue with Azure.” “We were able to build the Document AI solution using…Jun 5, 20245Document ProcessingAzure AI Document Intelligence · Azure Cognitive Services · Microsoft Azure02024-06-055
One area that the bank needed to automate to enhance its service was document processing. “Initially, a customer would bring the documents to a branch or email them, then our branch employee would print, scan, and then send them to the back office to manually extract the information," Georgia Zevgaropoulou elaborates.
↗So, it was an easy decision to continue with Azure.” “We were able to build the Document AI solution using Azure in less than four months.” Georgia Zevgaropoulou, Head of AI Sector in Artificial Intelligence and Technology Innovation Division, National Bank of Greece NBG started by analyzing 50 million pages of documents to create a dataset.
↗The bank serves 6 million customers and operates 356 branches across Greece.
↗This type of information can be crucial for our reporting, regulatory, or compliance matters.” With over 700,000 pages processed per month, this Azure-powered solution has become integral to NBG’s banking services.
↗We needed to process documents from more than 2 million customers in our digital channels in real time, and we needed to be accurate and consistent,” Georgia Zevgaropoulou adds.
Reported impact
Business outcome
This cutting-edge solution has improved the bank’s accuracy to 90%. By seamlessly integrating the solution with customer-facing systems, NBG has transformed its customer experience, enabled a fully digital journey, and drastically reduced waiting times.
Compliance
Customer Experience
Solution
National Bank of Greece used Azure AI Document Intelligence, Azure Cognitive Services, Microsoft Dynamics 365 to turn complex business content into faster digital workflows. One area that the bank needed to automate to enhance its service was document processing. “Initially, a customer would bring the documents to a branch or email them, then our branch employee would print, scan, and then send them to the back office to manually extract the information," Georgia Zevgaropoulou elaborates. Key implementation elements included Azure AI Document Intelligence extracts and structures information from business documents; Azure Cognitive Services adds speech, vision, or language capabilities required by the experience; Microsoft Dynamics 365 provides the common business process layer that standardizes operations; Microsoft Azure hosts the application, data, and integrations on scalable cloud infrastructure. Business documents and enterprise content are ingested into the AI workflow,...
Data flow
Business documents and enterprise content are ingested into the AI workflow, classified or summarized by the model layer, and routed into the application or operations process where employees can review and act on the output.
Solution architecture
4 components · 2 layers
Compute
Azure AI Document IntelligenceExtracts and structures information from business documents.
Azure Cognitive ServicesAdds speech, vision, or language capabilities required by the experience.
Microsoft AzureHosts the application, data, and integrations on scalable cloud infrastructure.
Orchestration
Microsoft Dynamics 365Provides the common business process layer that standardizes operations.
Architecture clues
Operates with real-time or near-real-time telemetry and analytics.
Evidence from the source
One area that the bank needed to automate to enhance its service was document processing. “Initially, a customer would bring the documents to a branch or email them, then our branch employee would print, scan, and then send them to the back office to manually extract the information," Georgia Zevgaropoulou elaborates.
So, it was an easy decision to continue with Azure.” “We were able to build the Document AI solution using Azure in less than four months.” Georgia Zevgaropoulou, Head of AI Sector in Artificial Intelligence and Technology Innovation Division, National Bank of Greece NBG started by analyzing 50 million pages of documents to create a dataset.
The bank serves 6 million customers and operates 356 branches across Greece.
This cutting-edge solution has improved the bank’s accuracy to 90%. By seamlessly integrating the solution with customer-facing systems, NBG has transformed its customer experience, enabled a fully digital journey, and drastically reduced waiting times.
“When we tried Azure AI Document Intelligence—or Form Recognizer, as it was known then—for document processing, the results were impressive in terms of accuracy and the time to market.
Review how this reference was scored and trace each normalized claim back to its documented source.
01
Scoring methodology
How scores are calculated
Customer Reference Value measures usefulness as a customer reference. Evidence Confidence measures the quality and traceability of available evidence, not independent verification.
Customer reference value65/100
Problem clarity20/20
Measures whether the customer's operational problem is specific enough to compare.
Implementation depth20/25
Rewards named components, data flow, integrations, and architecture detail.
Outcome specificity20/20
Measures whether the reported benefit is concrete rather than a generic claim.
Quantified impact0/25
0 metrics describe measurable impact.
Deployment scale5/10
Rewards evidence of a live or scaled deployment and meaningful integration scope.
Evidence confidence90%
Source authority20/25
The main evidence is an official vendor-published customer story.
Claim traceability23/25
11 of 12 key fields link to a named source.
Direct evidence20/20
5 direct source quotes retained.
Date provenance10/10
The publication date comes from the source page or publisher.
With your OK, we use Google Analytics cookies to see which pages and searches are useful. No ads, no profiling, and
nothing is loaded until you accept. You can change this any time under “Cookie settings” in the footer.