AArete's Doczy.ai Achieves 99% Contract Extraction Accuracy on AWS, Saving Healthcare Clients $330M

Title

AArete's Doczy.ai Achieves 99% Contract Extraction Accuracy on AWS, Saving Healthcare Clients $330M

Content

AArete's Doczy.ai contract interpretation platform, built end-to-end on Amazon Web Services, has processed 2.5 million contract documents (50 million pages) over 22 months, with 137 million API calls to Amazon Bedrock and 442 billion tokens processed. The platform delivers 99% extraction accuracy, a 97% reduction in manual processing time, and approximately $330 million in cumulative direct and indirect savings for healthcare and financial services clients. "Doczy.ai directly addresses these challenges using advanced AI and scalability on AWS," the post states. "Developed by AArete, it pushes the boundaries of document intelligence. The solution automatically interprets complex documents and converts them into a structured, queryable information repository that allows organizations to unlock the full value of their data and drive smarter decisions." For healthcare and financial services organizations, managing contracts, legal agreements, provider arrangements, and vendor invoices represents a major operational bottleneck. Manual review requires deploying teams to extract data from thousands of documents — error-prone and unsustainable. Existing Contract Lifecycle Management (CLM) systems prove inadequate for nuanced terms unique to each agreement, and reliance on institutional knowledge creates succession-planning risks. Prior to 2020, processing was manual at roughly 100 documents per week. From 2020 to 2023, AArete ran rules-based processing at about 55% accuracy. The breakthrough came in 2024 with AI processing on AWS, reaching 99% accuracy — nearly double the rules-based baseline. External users access the platform through a secure Next.js frontend with Amazon Cognito authentication, then upload documents to Amazon S3. AWS Lambda triggers Amazon Textract for text and metadata extraction. Doczy.ai's patented "smart chunking" algorithm preserves hierarchical structure via semantic and keyword search, decomposing text into context-aware chunks. A dual clustering engine analyzes every contract from both semantic (embeddings) and structural (pattern recognition for clause types, formatting) perspectives, with projection algorithms synthesizing them into a unified document model. Structured data flows into Snowflake; Amazon CloudWatch monitors performance; AWS Secrets Manager safeguards credentials. The platform scales to 250,000 contract documents per week, integrating with contract management systems like Coupa and Icertis. End-to-end automation feeds accurate contract data directly into claims systems, replacing manual translation of reimbursement terms, and continuously assesses payments against contract terms to flag potential overpayments or underpayments. AArete offers Doczy.ai as a SaaS solution, with AArete's experts configuring the platform for each client's document types, terminology, and workflows. The architecture continues extending into adjacent document-heavy processes across healthcare and financial services.