More Trackable Healthcare
Many patients struggle to keep a consistent medication schedule. At the same time, food-drug interactions are often overlooked, even though this information is important for helping patients stay careful.
Jivara is built as an AI-powered healthcare platform that helps patients manage medication schedules, monitor adherence, scan food, review potential food-drug interactions, and receive automatic reminders.
For healthcare workers, Jivara provides a monitoring dashboard. Nurses and admins can view patient progress, manage medication schedules, monitor adherence, and understand risks more quickly.
Benefits for Patients
For patients, Jivara provides a more guided experience. The dashboard displays active medications, daily schedules, medication status, and adherence history through a 12-month heatmap.
The Food Scan feature helps patients upload or scan food, then the system analyzes potential interactions with active medications. The result is not only a risk label, but also an explanation and recommendation that is easier to understand.
Because Jivara is built as a PWA, medication reminders can be sent through browser notifications. This makes the system closer to users' daily routines.
- Medication reminders help patients maintain their schedule.
- The adherence heatmap makes medication habits easier to review.
- Food Scan builds awareness of food and medication interactions.
- AI recommendations help users understand risk context more clearly.
Benefits for Healthcare Workers
Jivara does not only focus on patients. The platform also supports nurses and admins in monitoring patients remotely through a dashboard. Schedule data, medication status, and activity history can help healthcare workers identify patient adherence patterns.
When a nurse creates a patient, that patient can automatically be connected to the related nurse. This flow keeps monitoring clear and makes responsibility more structured.
With analytics, charts, and alerts, healthcare workers can identify patients who need attention more quickly.
Technology Behind the Product
Jivara uses a monorepo architecture with a Next.js 16 frontend and an Express.js 5 backend. The frontend is built with React 19, Tailwind CSS 4, Zustand, Chart.js, a service worker, and a PWA manifest. The backend uses TypeScript, PostgreSQL or Supabase, Drizzle ORM, JWT, Web Push VAPID, and API documentation through Scalar.
Food Scan runs through several steps: the image is uploaded, the backend creates a record, the food detection service reads the image, then the reasoning service checks potential interactions between the food and the patient's active medication. OpenRouter is used to generate interaction explanations and overall recommendations stored as snapshots.
Snapshots matter because analysis results can be reopened without repeating the AI process from the beginning. This makes the user experience more stable and makes results easier to audit.
- Next.js PWA makes the application more comfortable to use on users' devices.
- Express and Drizzle keep the backend structured.
- Web Push VAPID supports medication reminders.
- OpenRouter helps generate more communicative AI explanations.
Project Advantages
Jivara's strength lies in combining daily-use features with relevant AI. Medication reminders solve routine problems, adherence monitoring supports evaluation, while Food Scan adds value through food and medication awareness.
The project is also designed responsibly. In healthcare, AI must not appear to replace medical professionals. That is why analysis results need clear context, status, and limitations.
With patient and healthcare worker flows in one platform, Jivara has the potential to become a companion system that makes communication and monitoring more organized.