Modernizing Healthcare with Generative AI using Deepgram on AWS
Introduction
The world of healthcare is moving faster than ever, and GenAI is now the engine driving how clinicians and patients connect, document, and act on medical data. Deepgram, together with AWS Bedrock, puts medical-grade voice AI at the heart of this transformation. Nova-3 Medical leads the industry by delivering highly accurate, real-time transcription (even in noisy clinics!) outperforming competitors with a 63.7% improvement in word error rate and a 40% boost in capturing critical medical terms. Built for flexibility, it offers near-instant responses, HIPAA-grade security, and easy ways to adapt to new drugs, specialties, or workflows, all while scaling smoothly with your needs. Deepgram doesn’t provide an out-of-the-box app; instead, you get a set of powerful APIs and developer tools, the voice intelligence stack underneath, so your team can build the clinician and patient-facing experiences you need, while Deepgram handles the complex voice technology behind the scenes
In this blog post, we'll explore how Deepgram Nova-3 Medical, combined with Amazon Bedrock and Amazon Elastic Kubernetes Service (Amazon EKS), creates a powerful solution for addressing critical scaling challenges in healthcare operations. This integration offers real-time, HIPAA-compliant capabilities that improves patient care and streamline clinical workflows.
Solution Overview
Deepgram Nova-3 Medical
Deepgram Nova-3 Medical is an advanced speech-to-text AI model specifically trained on healthcare terminology. It boasts industry-leading accuracy, achieving the lowest Word Error Rate (WER) in clinical settings.
Amazon Bedrock
Amazon Bedrock is a fully managed service that provides easy access to high-performing foundation models (FMs) from leading AI companies. It offers a unified API, advanced security and privacy controls, and customization options, enabling developers to build and scale generative AI applications quickly and securely.
Amazon Elastic Kubernetes Service (Amazon EKS)
Amazon EKS is a managed Kubernetes service that simplifies the deployment, management, and scaling of containerized applications. It provides a highly available and secure environment for running Kubernetes workloads, integrating seamlessly with other AWS services.
By integrating Deepgram Nova-3 Medical with Amazon Bedrock for GenAI orchestration and Amazon EKS for scalable deployment, we create a robust platform for healthcare solutions.
Key Benefits:
Industry-leading accuracy in speech recognition for medical terminology
Real-time transcription delivering with low latency (<200ms)
HIPAA-compliant infrastructure
Scalable and flexible deployment options
Seamless integration with existing healthcare IT systems
Key use cases in healthcare
Healthcare organizations adopt this integrated solution across critical workflows, transforming traditional processes into efficient, AI-powered operations. Here are four use cases that show immediate value in clinical settings:
AI-Powered medical scribes: Streamline clinical documentation by transcribing doctor-patient conversations in real-time, auto-generating clinical notes with precision. This reduces the administrative burden on healthcare professionals, allowing them to focus more directly on patient care.
Automated drug prescriptions and refills: Enable automated interpretation of physician orders. This facilitates seamless integration with pharmacy management systems for timely and accurate medication dispatching, reducing the likelihood of human error.
Enhanced shipping and logistics: Improve medication dispatching and supply chain management through automated interpretation of orders and integration with logistics systems.
Intelligent registration and scheduling: Automate patient registration and appointment scheduling, cancellations, and follow-ups through conversational AI, improving efficiency and patient experience.
Architecture
To operationalize these powerful GenAI-driven applications, Deepgram partners closely with AWS to ensure seamless and secure deployment. Two deployment options are available:
Customer-owned deployment: In this approach deploy Deepgram Nova-3 Medical directly within customer-managed AWS accounts. This ensures all customers data remains within their own AWS account, providing full control and seamless integration with existing healthcare IT infrastructures.
Deepgram platform is deployed on an Amazon EKS cluster, hosting inside the customer’s Amazon Virtual Private Cloud (VPC). Customers access the platform through AWS API Gateway, which securely routes requests to the EKS cluster. Amazon Simple Storage Service (Amazon S3) and Amazon Elastic Container Registry (Amazon ECR) provide essential storage and container management capabilities.
In the customer's AWS account, Amazon Bedrock orchestrates foundation models, account lookup and booking processes. AWS Lambda functions integrate with a Knowledge Base and Electronic Health Record FHIR API endpoint. This serverless architecture ensures scalable, secure processing of healthcare data workloads.
Software as a Service (SaaS) deployment
Deepgram also offers a dedicated SaaS implementation option for customers prefer to leverage a Deepgram-managed environment to simplify their day-to-day operations and ensures HIPAA compliance and data security. Client applications connect to the AWS API Gateway and Amazon EKS cluster sitting inside Deepgram’s AWS account and VPC. Similar to the customer-owned deployment, it also uses Amazon S3 for voice and transcript storage, and Amazon ECR for container management.
The knowledge assets, customer application stack, and tools will be hosted within the customer’s AWS account. Amazon Bedrock manages integrations through AWS Lambda functions. The system interfaces with a custom Knowledge Base and Electronic Health Record FHIR API for healthcare data processing.
Both deployment options are designed to be highly scalable and flexible by utilizing Amazon EKS for container orchestration and Amazon Bedrock for AI model management. These AWS services ensure the solution can scale with the needs of healthcare organizations of any size.
User Experience
To demonstrate the capabilities of Deepgram Nova-3 Medical on AWS, let's explore an example use case that integrates Amazon Bedrock's function-calling LLM with Deepgram's Voice Agent API.
Audio data from doctor-patient conversations is streamed to Deepgram Nova-3 Medical using the Voice Agent API.
The transcribed text is then processed using Amazon Bedrock's function-calling LLM to interpret user intents and generate clinical notes or perform other tasks.
The output is stored in Amazon S3 for further processing and integration with other healthcare systems.
Deepgram Powers the Brains, You Build the Experience
One thing to be clear about: Deepgram is not a ready-made healthcare app. Instead, it’s the intelligence stack behind the scenes. Think of Deepgram as the engine and the toolkit, powerful APIs and models that handle the complex process of turning voice into meaning, transcribing, analyzing, and orchestrating information.
Healthcare developers, IT teams, digital health innovators create the experiences and user interfaces that fit your clinicians, your workflows, and your patients. Whether you’re building a digital scribe, automating pharmacy refills, or designing entirely new experiences, Deepgram gives you the building blocks. You bring the vision, Deepgram provides real-time voice intelligence to make it possible.
Healthcare’s demands will only grow. Deepgram is ready so your teams can spend less time wrangling paperwork and more time focused on care.
Conclusion
The combination of Deepgram’s specialized medical AI capabilities with AWS’s robust infrastructure creates a foundation for sustainable digital transformation in healthcare. Integrating Deepgram Nova-3 Medical with AWS Bedrock and Amazon EKS delivers concrete benefits for healthcare organizations embracing Generative AI. Through this solution, healthcare organizations achieve:
Industry-leading accuracy in medical terminology transcription with <200ms latency
Automated clinical documentation that reduces administrative workload
Streamlined prescription management and pharmacy integration
Enhanced patient registration and scheduling workflows
The flexible deployment options cater to diverse organizational needs, delivering HIPAA-compliant operations. Customer-Owned deployment offers control over data, usage and within customer’s AWS account, ideal for large customers with strict data sovereignty requirements. SaaS deployment provides a Deepgram-managed, HIPAA-compliant environment for organizations seeking simplified operations and reduced maintenance overhead.
Both options use Amazon EKS for scalable container orchestration and Amazon Bedrock for secure AI model management. This helps healthcare customers scale their operations while maintaining compliance and security standards.
Call to Action
Ready to experience voice-to-text technology? Visit the Deepgram playground and sign up for a free account to test drive it using your own voice. When you are ready to develop your solution, subscribe to Deepgram Nova-3 Medical via the AWS Marketplace. A free trial is available to help you experience firsthand how this solution can revolutionize your clinical operations and patient engagement.
Learn More
Amazon Bedrock function-calling schema and proxy URL
Unlock language AI at scale with an API call.
Get conversational intelligence with transcription and understanding on the world's best speech AI platform.