Normalized claim
Document processing efficiency: 400% increase
"This increased our document processing efficiency by 400%"
GCPEvidence: Low35/100
Mr. Cooper is using AI to increase speed and accuracy for mortgage processing
Use case typeDocument processing automationUpdated Jun 4, 2021
Mr. Cooper Group built a highly reliable, cloud-native document analysis and processing platform to process lending documents. The platform was designed to improve speed, accuracy, and cost control for mortgage servicing and origination workflows while supporting a better digital experience for homeowners.
- Organization
- Mr. Cooper Group
- Industry
- Finance
- Location
- United States
- Published
- June 2021
Reported outcomes
+400%
document processing efficiencyProductivity & throughput
+95%critical document accuracy4,000 pages/minpeak throughput2,000 pages/minaverage throughput
Strategic outcomes
Cost efficiencyReduced platform costsCustomer experience & trustImproved digital customer experienceSpeed & agilityShortened loan servicing time
Catalog median for productivity & throughput deployments: +40% across 108 reported metrics. Compare benchmarks →
Why do we believe this?Outcome claims, sources, and evidence checks
Normalized claim
Critical document accuracy: 95% increase
"accuracy of over 95% for critical documents"
Normalized claim
Peak throughput: 4,000 pages/min increase
"a peak throughput of 4000 pages/min"
Normalized claim
Average throughput: 2,000 pages/min increase
"an average throughput of 2000 pages/min"
Why do we believe this deployment?Customer identity, provider attribution, maturity, and source checks
- Customer
- Mr. Cooper Group
- Provider
- GCP
- Maturity
- Unknown
- Linked source
- Google Cloud Blog
No explicit deployment-stage evidence found.
Customer identity supportedSource describes one deploymentMaturity evidence evaluated
Primary read
Use case focus
Showing 2 of 2
- 1Document processing automation
- 2Operations optimization
- Reduce mortgage document processing time and cost beyond typical OCR.
- Accurately identify, classify, and extract document data at high throughput while improving the digital customer experience.
- Built a modular, container-based, API-first document pipeline on Google Cloud.
- Used Document AI and Vertex AI to manage mortgage-specific model versions for classification and extraction.
- Processed documents through asynchronous event handling on Google Kubernetes Engine and exposed document data through Apigee APIs.
- Used BigQuery for analytics and Cloud SQL for database management, while retraining models as document formats and data drift changed.
- Achieved over 95% accuracy for critical documents.
- Reached peak throughput of 4,000 pages per minute and average throughput of 2,000 pages per minute.
- Increased document processing efficiency by 400%.
- Reduced platform costs significantly.
Architecture
A modular, container-based, API-first document processing platform on Google Kubernetes Engine with Document AI and Vertex AI for mortgage-document classification and extraction, Apigee for API exposure, BigQuery for analytics, and Cloud SQL for database management. The workflow uses asynchronous event processing and ongoing model retraining to handle document drift and changing formats.
Sources & evidence1
Evidence: Low35/100Evidence strength
- Customer explicitly identified
- Quantified outcome available
- Technical implementation details available
Type: Blog PostPublished: Jun 4, 2021Publisher: Google CloudEvidence: VendorConfidence: High
AI-generated summary. Verify important details with the linked sources before relying on this case.
Explore related AI use cases
Customer evidence
Provider evidence
Was this useful?
Community
Comments
No published comments yet.