Swasthya Sathi

Multilingual AI health guidance for faster, safer decisions in underserved communities.

Swasthya Sathi is a mobile-first multilingual health assistant designed to help people understand symptoms, gauge urgency, and get simple next steps in their preferred language. It uses a guided triage flow, plain-language guidance cards, and nearby resource recommendations to support users in deciding whether to rest, visit a clinic soon, or seek urgent care.

Inspiration

We built Swasthya Sathi because getting timely, understandable health guidance is still hard for too many people, especially when language, literacy, and access barriers get in the way. We wanted to create something that feels calm, clear, and useful in the moment someone is worried about symptoms, instead of overwhelming them with medical jargon.

What it does

Swasthya Sathi is a mobile-first multilingual health assistant that helps users understand symptoms, gauge urgency, and get simple next steps in their preferred language. It uses a guided triage flow, plain-language guidance cards, and nearby resource recommendations to help people decide whether to rest, visit a clinic soon, or seek urgent care. It also includes reminders, a low-literacy mode, and an admin dashboard for NGOs or clinics to understand community needs better.

How we built it

We built the app with Next.js 14, TypeScript, Tailwind CSS, and shadcn/ui to keep the experience fast and mobile-friendly. Prisma and SQLite power the local data layer for reminders, curated resources, and lightweight analytics, while Zod and React Hook Form help keep the triage flow structured and safe. We also designed the multilingual layer with next-intl or a custom i18n approach, and made the deployment Vercel-compatible with Docker support for portability.

Challenges

One of the biggest challenges was balancing helpfulness with safety, since health guidance needs to be clear without pretending to diagnose. We also had to design the triage flow so it stayed simple enough for low-literacy users while still collecting enough information to make the guidance meaningful. Building a multilingual experience that feels natural across different languages and screen sizes took a lot of iteration.

Accomplishments

We’re proud that the app feels genuinely usable for a real person in a stressful moment, not just like a hackathon demo. The safety-first triage flow, plain-language guidance, and low-literacy mode came together into something that can actually support underserved communities. We also like that the project has a clear path to real-world use through NGO and clinic workflows.

What we learned

We learned how much thoughtful UX matters when building for health, especially when users may be anxious, unfamiliar with medical terms, or using the app on a low-end phone. We also learned how to structure a product around safety constraints instead of treating them as an afterthought. On the technical side, we got better at building modular Next.js apps, handling multilingual content cleanly, and designing data models that support future growth.

What's next

Next, we want to improve the triage engine with more localized guidance, richer language support, and better accessibility features like voice input and text-to-speech. We’d also love to connect the resource directory to real NGO and clinic partners, and expand the admin dashboard into a tool that helps communities spot care gaps over time. Longer term, we want Swasthya Sathi to become a trusted first step for people who need health guidance but don’t know where to start.

Built With

• docker
• lucide-react
• next-intl-/-custom-i18n
• next.js-14-(app-router)
• optional-framer-motion
• optional-openai-compatible-local-mock-llm-layer
• optional-recharts
• prisma-orm
• react-hook-form
• shadcn/ui
• sqlite
• tailwind-css
• typescript
• vercel-compatible-deployment
• zod