AI-Driven Healthcare Document Automation for Gabby
Gabby, a healthcare technology provider, needed to handle large volumes of medical and operational documents securely and accurately. Manual document processing was time-consuming, error-prone, and costly to scale.
Challenges
- Healthcare organizations face strict data governance requirements and the operational burden of maintaining consistent accuracy across diverse document formats. Gabby’s existing workflows relied heavily on manual review, limiting throughput and raising compliance risks.
The Solution
KnackForge implemented an AI-enabled document intelligence platform using AWS-native services to transform unstructured healthcare data into query-ready insights.
Amazon Textract extracts structured data from scanned medical records and uploaded documents.
AWS Bedrock (Claude 3.5) and LangChain power summarization, predictions, and conversational analytics.
AWS Lambda for serverless automation of data synchronization, workflow execution, and background AI tasks.
Django API on Amazon ECS manages multi-tenant workflows and integrations.
MatrixCare API syncs resident data at scheduled 5-minute intervals.
Data and embeddings are stored in Amazon RDS (PostgreSQL + PGVector), with React frontends delivered through CloudFront and S3. The solution leverages VPC isolation, AWS WAF, Secrets Manager, and CloudWatch for security, monitoring, and compliance.
The Impact
- 90% reduction in manual document handling.
- Near real-time data availability for analytics and claims workflows.
- Higher accuracy and audit-ready compliance posture.
Gabby’s document workflows are now fully automated, transforming static text into actionable, structured healthcare data.
Technologies Used:
- Amazon Textract
- Amazon Comprehend Medical
- AWS Lambda
- Amazon S3