🩺 Akan Health Assistant

🌍 Inspiration

Akan Health Assistant was inspired by the language barrier many people face when accessing digital healthcare tools. Most AI health platforms operate only in English, limiting accessibility for local language speakers. I wanted to build an inclusive AI health assistant that supports both English and Akan, while providing an engaging, human-like experience.

🚀 What I Built

A full-stack AI-powered health assistant featuring:

• 🤖 3D AI Doctor Interface using React Three Fiber with speech synthesis
• 🌍 Multilingual Support (English & Akan)
• 🔍 Real-Time Web Search Integration via DuckDuckGo API
• 📊 Health Records & Analytics Dashboard
• 🔐 JWT-Based Authentication System
• 📱 Fully responsive UI

🛠 Tech Stack

Frontend: Next.js 14, TypeScript, Tailwind CSS, shadcn/ui Backend: FastAPI, MongoDB, Motor, Pydantic, JWT Search Integration: aiohttp + DuckDuckGo

The system processes a health query by detecting language, querying a local medical knowledge base, optionally enhancing results with real-time web data, and returning a confidence-scored response:

[ Confidence = \frac{Relevance + ContextMatch + SourceReliability}{3} ]

🧠 Key Challenges

• Handling multilingual medical terminology in Akan
• Optimizing 3D rendering performance across devices
• Managing async operations (MongoDB + web search) efficiently
• Securing sensitive health data with proper authentication and validation

📚 What I Learned

This project strengthened my skills in:

• Full-stack architecture design
• Async backend development with FastAPI
• Type-safe frontend development with TypeScript
• 3D web rendering and performance optimization
• Building inclusive, socially impactful AI systems

🌟 Impact

Akan Health Assistant demonstrates how AI can bridge language gaps in healthcare, combining accessibility, real-time intelligence, and immersive interaction into a scalable digital health solution.

Built With

• fastapi
• mongodb
• nextjs
• typescript