Here is the improved English version formatted cleanly in Markdown:

🩺 About ClinicLite

ClinicLite is an electronic health record (EHR) system designed specifically for the realities of rural West Africa, especially Senegal. It works fully offline, allowing health posts, community clinics, and remote medical centers to manage patient information even without an Internet connection. With integrated local AI, healthcare workers get support for documentation, summaries, and basic clinical tasks — all running on-device, without the cloud.

🌱 Why ClinicLite?

Across Senegal and neighboring countries, many health facilities still face:

• Frequent Internet disruptions
• Old or low-power computers
• Limited access to modern digital health tools
• Heavy reliance on paper records

ClinicLite was built to address these challenges with a simple principle: A lightweight, resilient tool that works anywhere — even in the most isolated communities.

🛠️ Technology Built for Local Realities

We selected open-source technologies that perform well on modest hardware:

• ⚙️ Backend: Node.js + Express — fast, stable, resource-efficient
• 🖥️ Frontend: React (Vite) PWA — fully offline via PouchDB + service workers
• 🧠 Local AI: TinyLLaMA via Ollama — clinical assistance without Internet
• 💾 Databases: SQLite (server) & PouchDB (browser) — lightweight and reliable
• 📦 Deployment: Docker Compose — easy installation on any PC or LAN
• 🛜 Optional Sync: Secure data synchronization when connectivity becomes available

🔁 How It Works in the Field

• The PWA runs directly in the browser and does not require Internet
• Patient records are stored locally for constant availability
• The backend automatically syncs data when connection is restored
• Local AI assists with summaries, note-taking, and treatment suggestions — even in areas without mobile network coverage
• The entire system can be deployed in a district health office or rural clinic through a simple LAN

🚧 Challenges Overcome

• Running an AI model on low-spec hardware with no GPU
• Achieving smooth PouchDB ↔ SQLite synchronization in offline-first mode
• Ensuring secure authentication and AI inference work reliably offline
• Packaging everything into a deploy-anywhere Docker setup for local IT teams

💡 What We Learned

• Modern lightweight AI makes it possible to deliver real assistance in rural clinics
• Offline-first design is not a fallback — it’s a core strength for resilience and equity
• With TinyLLaMA and Ollama, AI becomes accessible without cloud services, subscriptions, or expensive infrastructure

Built With

• docker
• express.js
• ollama
• react
• sqlite