Mimamori - AI Health Companion
Mimamori (meaning 'watching over' in Japanese) was built to transform how chronic conditions and daily wellness are monitored. Developed using AWS Bedrock and Kiro IDE, the platform ensures that the 99.9% of time patients spend outside of clinics is no longer a data 'black box.' Patients spend over 8,700 hours a year outside clinical settings—hours where vital health data is lost. Mimamori solves this by providing a continuous logging mechanism via voice, enabling seamless updates to caregivers through a Shared Care Circle Dashboard, and extracting unstructured data into Doctor-Ready reports using medical-grade NLP.
8
Tech Stack
6
Key Features
Technologies Used
Next.js 14
React 18
TypeScript
AWS Bedrock (Nova)
Amazon Comprehend Medical
Amazon Textract
DynamoDB
Prisma
Key Features
- ●Voice Symptoms Log: AI-powered voice capture that understands colloquial nuances and non-linear speech.
- ●Clinical Synthesis: Automated extraction of medications, conditions, dosages, and vitals using medical-grade NLP.
- ●Doctor-Ready Reports: Comprehensive PDF health summaries with longitudinal trend visualizations and anomaly detection.
- ●Smart Alerts: Real-time notifications for pulse/oxygen anomalies or worsening symptom trends using SNS/SES.
- ●Health Vault: Secure, encrypted storage for lab reports and medical prescriptions with OCR parsing via Textract.
- ●Care Circle: Granular permission settings to share health updates with family and medical teams.