What challenges did Clarus Care face before AI implementation?

High volumes of patient calls create long hold times and overwhelm staff who manually prioritize and process messages. A traditional menu-driven IVR cannot handle complex, multi-intent patient needs or provide natural conversational interactions efficiently.

What was the impact of the AI solution?

The article says the solution improves patient communication, reduces staff workload, and lowers hold times. Clarus serves over 16,000 users across 40+ specialties and handles 15 million patient calls annually. The prototype targets a 99.99% SLA and sub-3-second backend latency, indicating a production-oriented design for high-volume patient communications.

Source

Evidence: Low35/100

How Clarus Care uses Amazon Bedrock to deliver conversational contact center interactions

Use case typeCustomer support automationUpdated Jun 13, 2026

Clarus Care is a healthcare technology company that helps medical practices manage patient communication through an AI-powered call management system. The company is building a generative AI-powered contact center prototype to support conversational voice and chat interactions across appointments, prescription refills, billing inquiries, and urgent medical concerns. The solution uses Amazon Connect, Amazon Lex, and Amazon Bedrock with Claude and Amazon Nova models to transcribe and route calls, detect urgency, extract multiple intents in one interaction, support smart transfers, and provide analytics.

Organization: Clarus Care
Industry: Healthcare
Location: United States
Published: February 2026

Reported outcomes

100%

quantified impactTime & speed

Strategic outcomes

Customer experience & trustImproved patient communicationCost efficiencyReduced staff workloadSpeed & agilityLowered hold timesNew product / capabilityBuilt conversational voice and chat support

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

Normalized claim

Quantified impact: 100%

AWS Machine Learning BlogFeb 2, 2026Blog postInferred claimLow evidence strength

The prototype targets a 99.99% SLA and sub-3-second backend latency, indicating a production-oriented design for high-volume patient communications.

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

Customer: Clarus Care
Provider: AWS
Maturity: Unknown
Linked source: AWS Machine Learning Blog

No explicit deployment-stage evidence found.

Customer identity supportedSource describes one deploymentMaturity evidence evaluated

Primary read

Use case focus

Showing 3 of 3

1Contact Center Automation
2Conversational AI
3Customer Support

High volumes of patient calls create long hold times and overwhelm staff who manually prioritize and process messages.
A traditional menu-driven IVR cannot handle complex, multi-intent patient needs or provide natural conversational interactions efficiently.

Built a multichannel contact center prototype on Amazon Connect and Amazon Lex.
Used Amazon Bedrock with Anthropic Claude 3.5 Sonnet and Amazon Nova models through the Converse API for urgency assessment, multi-intent detection, information extraction, and natural response generation.
Implemented smart transfer handling for urgent cases and provider requests, plus a scheduling module, conversation-state tracking, and an analytics pipeline with dashboarding.
Added a web chat interface to demonstrate the same service experience across voice and chat channels.

Technologies

Amazon Connect Amazon Lex Amazon Bedrock Anthropic Claude 3.5 Sonnet Amazon Nova Pro Amazon Nova Lite Bedrock Converse API Amazon Bedrock Knowledge Bases

The article says the solution improves patient communication, reduces staff workload, and lowers hold times.
Clarus serves over 16,000 users across 40+ specialties and handles 15 million patient calls annually.
The prototype targets a 99.99% SLA and sub-3-second backend latency, indicating a production-oriented design for high-volume patient communications.

Architecture

Amazon Connect handles voice and chat entry points and routes contacts through configured contact flows. Amazon Lex manages transcription and session attributes. A fulfillment function invokes Amazon Bedrock models, including Anthropic Claude 3.5 Sonnet and Amazon Nova Pro/Lite, via the Converse API to classify urgency, extract multiple intents, gather missing information, and generate responses. Urgent or provider-request cases are handed off to staff. A separate scheduling module handles appointment workflows, and conversation logs are processed into an analytics pipeline and dashboard. The post also mentions Bedrock Knowledge Bases and content/PII/PHI safeguards.

Sources & evidence1

Evidence: Low35/100Evidence strength