Jaystring20/bio-guide-naija

# VeriDIA - AI Health Intelligence Platform > An agentic AI platform that interprets Nigerian medical lab results and generates culturally-grounded, personalized dietary recommendations for NCD management. ## 🎯 The Problem Nigeria has a doctor-to-patient ratio of 1:3,474. Millions of Nigerians receive medical lab results with no clinical interpretation, leaving them confused and unable to act on critical health information. With 22.9% of adults affected by hypertension and ~4.7 million living with diabetes (75%+ undiagnosed or unmanaged), the gap between lab result and actionable health guidance is costing lives. VeriDIA exists to close that gap. ## 💡 The Solution VeriDIA is a **mobile-first AI health companion** that transforms medical lab results into actionable health intelligence through four core functions: 1. **Lab Result Interpretation** - Multimodal AI (OCR + NLP) extracts and interprets lab data from uploaded images/PDFs 2. **Plain-Language Explanation** - Translates clinical jargon into accessible Nigerian English 3. **Personalized Dietary Plans** - Context-aware recommendations using Nigerian foods, cooking methods, and cultural eating patterns 4. **Doctor's Checklist** - Structured summary for physician consultation with flagged abnormalities ## 🏗️ Architecture ### Tech Stack **Frontend:** - React 18+ with TypeScript - Vite (build tooling) - TailwindCSS (styling) - Shadcn/ui (component library) **Backend & Infrastructure:** - Supabase (PostgreSQL database, authentication, edge functions) - Google Gemini API (vision for OCR and clinical reasoning) - LLM integration for medical interpretation - Retrieval-Augmented Generation (RAG) for clinical verification **AI/ML Pipeline:** 1. **Perception Layer** - Gemini vision models extract structured data from lab documents 2. **Reasoning Layer** - LLM interprets clinical values against Nigerian reference ranges 3. **Verification Layer** - RAG cross-references against validated clinical databases …