ArcaAI
ArcaAI is an AI-powered electronic health record platform for clinicians, clinics, and hospitals, with ambient listening, prescription, lab, and clinical documentation modules. Kiebot built the backend services that turn doctor-patient conversations into structured clinical documents and keep patient data handled to healthcare standards across two clouds.
The Challenge
Ambient clinical documentation has to be accurate, fast, and private. The platform needed event-driven services that could run on both AWS and Azure, transcribe consultations into structured notes with diagnosis coding, mask and encrypt personal health information at the field level, and give the team visibility into how every AI prompt performed in production.
Our Solution
Kiebot delivered a tailored solution focusing on scalability and user experience. Key features include:
- Serverless, event-driven microservices on AWS Lambda and Azure Functions for healthcare data processing across both clouds.
- AWS Bedrock integration for ambient transcription of doctor-patient conversations into structured clinical documents.
- HIPAA-aligned data pipelines with PII masking and field-level encryption, including ICD-11 diagnosis coding and structured clinical notes.
- Langfuse for LLM observability and prompt performance tracking, so clinical output quality is measured, not assumed.
- High-throughput NoSQL data models across DynamoDB and CosmosDB, with OpenTelemetry distributed tracing across services.
Tech Stack
Key Results
- AWS + AzureMulti-cloud serverless backend
- HIPAAAligned data pipelines
- ICD-11Coded clinical notes