Logo Lanfrica
  • Home
  • Atlas
  • Insights
  • Docs
  • Sign in

© 2026 Lanfrica. All rights reserved. All copyrights of the resources shown on the Lanfrica website belong to the original copyright holders, unless explicitly stated otherwise.

DataHERAfrica/datathon-2025-HealthPaddie

Domain:

healthcarenatural language processing

Record type:

software
Creator:
Dat
Host:
HealthPaddie is an AI-powered health assistant designed to deliver trusted, accessible, and easy-to-understand health information, especially for underserved and rural communities in Africa. # datathon-2025-HealthPaddie // Team name- Data Girls :herb: HealthPaddie A RAG-based digital companion delivering verified health information in local African languages. ## Overview HealthPaddie is an AI-powered health assistant designed to provide trusted, verified, and easy-to-understand health information especially for people in rural or underserved communities across Africa. Built during the DataHER Africa Hackathon 2025 (#DHAD25), HealthPaddie helps users ask health-related questions and receive answers grounded in real medical documents (WHO, NCDC, UNICEF). HealthPaddie stands out with: Multilingual support — English, Hausa, Yoruba, Igbo Verified medical information — No hallucinations Streamlit conversational interface Text-to-Speech (TTS) for accessibility Groq LLM inference for fast, safe responses ## Problem Statement Access to verified medical information remains a challenge for many communities due to: Widespread misinformation Language barriers Lack of trusted sources Limited literacy HealthPaddie provides verified, multilingual health answers in a simple, accessible way. ## Key Features Document-Grounded Health Answers (RAG) Retrieves facts from a FAISS vectorstore built with curated medical datasets. Multilingual Support: Responses available in: - English - Hausa - Yoruba - Igbo AI-Powered Text-to-Speech Enables users with low literacy or visual challenges to hear answers in audio. Fast AI Response (Groq LLaMA 3.1) Ultra-low latency responses using Groq’s LLM API. ## Safety-First Approach No diagnoses Encourages medical consultation Detects potentially serious symptoms ## Built From WHO fact sheets American Society for Microsoft National Minority AIDS Council (NMAC) Health education PDFs Public medical datasets # RAG Pipeline Diagram ┌────────────────────────┐ │ User Input │ │ (Health question in │ │ English/Hausa/Yoruba/ │ │ Igbo) │ └────────────┬───────────┘ │ ▼ ┌────────────────────────┐ …

Visit

github.com

Languages

HausaIgboYoruba

Similar

DataHERAfrica/datathon-2025-hub

DataHERAfrica/datathon-2025-hub

The DataHER Africa Datathon 2025 brought together brilliant African women from across the continent