What challenges did Redox face before AI implementation?

The health system manually processed tens of millions of incoming faxes and expected volume to rise by more than 350% after an EHR rollout. Meeting the increased workload would have required about 4x staffing and could delay patient care. Redox needed to automate triage and routing of time-sensitive referral and prescription faxes while preserving HIPAA compliance.

What was the impact of the AI solution?

The solution achieved 85–90% document classification accuracy at launch. Redox went from a fully manual process to robust automation in 4 weeks. Fax processing time dropped from days to seconds. The automation reduced manual workload and operational cost while expediting patient care.

Source

Scaled productionEvidence: Medium65/100

Redox automates healthcare fax processing with Amazon Bedrock and Amazon Nova

Use case typePatient engagementUpdated Jun 13, 2026

Redox, a healthcare data interoperability platform, helped a major US health system automate the manual processing of patient faxes. The customer’s HIM team needed to keep up with a large and growing fax volume while maintaining HIPAA-compliant handling and routing documents to the correct downstream destinations.

Organization: Redox
Industry: Healthcare
Location: United States
Published: April 2026

Reported outcomes

85-90%

accuracyQuality & accuracy

Strategic outcomes

Speed & agilityAutomated fax processing in weeksNew product / capabilityAutomated document triage and routingCustomer experience & trustExpedited patient careCost efficiencyReduced manual workload and operational cost

Why do we believe this?Outcome claims, sources, and evidence checks

Normalized claim

Accuracy: 85-90%

AWS Customer StoryApr 29, 2026Customer storyInferred claimMedium evidence strength

The solution achieved 85–90% document classification accuracy at launch.

Why do we believe this deployment?Customer identity, provider attribution, maturity, and source checks

Customer: Redox
Provider: AWS
Maturity: Scaled Production
Linked source: AWS Customer Story

Redox iterated on prompt optimization and routing logic to improve classification accuracy and consistency at scale

Customer identity supportedSource describes one deploymentMaturity supported

Primary read

Use case focus

Showing 3 of 4

1Document processing
2Workflow automation
3Classification

The health system manually processed tens of millions of incoming faxes and expected volume to rise by more than 350% after an EHR rollout.
Meeting the increased workload would have required about 4x staffing and could delay patient care.
Redox needed to automate triage and routing of time-sensitive referral and prescription faxes while preserving HIPAA compliance.

Redox embedded generative AI into its orchestration layer using Amazon Bedrock.
The solution uses Amazon Nova Lite to classify incoming documents and route them to more than 1,200 destinations.
Redox iterated on prompt optimization and routing logic to improve classification accuracy and consistency at scale.
The system runs on Redox’s existing secure AWS infrastructure and was deployed as a modular addition rather than a full platform rebuild.

Technologies

Amazon Bedrock Amazon Nova Lite Amazon Nova

The solution achieved 85–90% document classification accuracy at launch.
Redox went from a fully manual process to robust automation in 4 weeks.
Fax processing time dropped from days to seconds.
The automation reduced manual workload and operational cost while expediting patient care.

Architecture

Redox embedded Amazon Bedrock into its existing orchestration layer and used Amazon Nova Lite for document classification. The AI capability was added as a modular step in a HIPAA-compliant AWS platform to classify incoming faxes and route them to downstream systems.

Sources & evidence1

Evidence: Medium65/100Evidence strength