🏥 Neural RX - AI-Powered Healthcare Search

Winner Submission for the Elastic + Google Cloud Hackathon Challenge
Combining Elasticsearch hybrid search with Google Gemini AI for intelligent healthcare information discovery

🌟 Overview

Neural RX is an intelligent healthcare information discovery platform that makes finding medical information feel like having a conversation with a knowledgeable medical reference guide. By combining Elasticsearch's hybrid search capabilities with Google Gemini's generative AI, we've created an experience that's both precise and personal.

✨ Key Features

🤖 Conversational AI Interface - Ask health questions in plain language
🔍 Hybrid Search - Combines keyword + semantic matching across 12+ medical conditions
🗣️ Voice Input - Hands-free interaction using Web Speech API
💾 Export Chat History - Download conversations in PDF, Text, or JSON format
⭐ Bookmarks - Star important messages for quick reference
🧠 Context-Aware Conversations - Remembers previous questions for natural follow-ups
⚡ Quick Actions - One-click access to 6 common health topics
🌓 Dark Mode - Beautiful UI that adapts to your preference
♿ Accessible - WCAG AA compliant with keyboard navigation

🎥 Demo

[Add screenshots or GIF demo here]

Try these sample queries:

"What are the symptoms of diabetes?"
"Tell me about treatment options for hypertension"
"How is asthma diagnosed?"
"What causes migraines?"

🏗️ Architecture

┌─────────────────────────────────────────────────────────┐
│                    User Interface                        │
│  ┌──────────────────────┐  ┌─────────────────────────┐ │
│  │   Chat Interface     │  │  Search Results Panel   │ │
│  │  - Voice Input       │  │  - Medical Cards        │ │
│  │  - Bookmarks         │  │  - Symptoms & Treatments│ │
│  │  - Export Options    │  │  - Severity Badges      │ │
│  └──────────────────────┘  └─────────────────────────┘ │
└─────────────────────────────────────────────────────────┘
                            │
                            ▼
┌─────────────────────────────────────────────────────────┐
│                   Express.js Backend                     │
│  ┌──────────────────┐  ┌──────────────────────────┐    │
│  │  Gemini AI       │  │  Elasticsearch           │    │
│  │  - Entity Extract│  │  - Hybrid Search         │    │
│  │  - Response Gen  │  │  - 12 Health Conditions  │    │
│  │  - Context Aware │  │  - Multi-field Matching  │    │
│  └──────────────────┘  └──────────────────────────┘    │
└─────────────────────────────────────────────────────────┘

Data Flow

User asks a health question via chat or voice
Gemini AI extracts medical entities from the query
Elasticsearch performs hybrid search across healthcare database
Search results provide context to Gemini AI
Gemini generates an informed, conversational response
UI displays both AI response and structured medical cards

🛠️ Technology Stack

Frontend

React 18 with TypeScript
Tailwind CSS + Shadcn UI for beautiful healthcare design
TanStack Query for data fetching and state management
Wouter for lightweight routing
Lucide React for icons

Backend

Express.js server with TypeScript
Elasticsearch (@elastic/elasticsearch) for hybrid search
Google Gemini AI (gemini-2.5-flash) for conversational intelligence
In-memory storage for chat history and analytics

Integrations

Elastic Cloud - Managed Elasticsearch deployment
Google AI Studio - Gemini API access
Web Speech API - Browser-native voice recognition

🚀 Getting Started

Prerequisites

Node.js 18+ and npm
Elastic Cloud account (Sign up free)
Google AI Studio API key (Get yours)

Installation

Clone the repository

git clone https://github.com/yourusername/neural-rx.git
cd neural-rx

Install dependencies
```
npm install
```

Set up environment variables

Create a .env file in the root directory:

GEMINI_API_KEY=your_gemini_api_key_here
ELASTIC_CLOUD_ID=your_elastic_cloud_id_here
ELASTIC_API_KEY=your_elastic_api_key_here
SESSION_SECRET=your_random_session_secret

Get Elasticsearch credentials
- Go to Elastic Cloud
- Create a new Deployment (not Serverless)
- Copy the Cloud ID
- Open Kibana → Stack Management → Security → API Keys
- Create an API key with full permissions
- Update ELASTIC_CLOUD_ID and ELASTIC_API_KEY in .env
Get Google Gemini API key
- Visit Google AI Studio
- Create a new API key
- Update GEMINI_API_KEY in .env
Run the application
```
npm run dev
```
Open your browser
```
http://localhost:5000
```

The backend will automatically:

Connect to Elasticsearch
Create the healthcare_conditions index
Seed 12 medical conditions
Start serving on port 5000

📊 Healthcare Database

Neural RX includes a curated database of 12 major medical conditions:

Conditions Covered

Endocrine: Diabetes Type 2, Hypothyroidism
Cardiovascular: Hypertension, Atrial Fibrillation
Respiratory: Asthma
Neurological: Migraine
Musculoskeletal: Osteoarthritis, Rheumatoid Arthritis
Mental Health: Depression, Generalized Anxiety Disorder
Gastrointestinal: GERD (Gastroesophageal Reflux Disease)
Renal: Chronic Kidney Disease

Each condition includes:

Detailed description
Common symptoms
Treatment options
Severity rating (Mild, Moderate, Severe)
Medical category classification

🎨 Design System

Neural RX follows a healthcare-focused design philosophy:

Color Palette: Medical blue primary (#0EA5E9) with carefully chosen accent colors
Typography: Inter for body, JetBrains Mono for technical data
Spacing: Consistent 2/4/6/8 unit system
Accessibility: WCAG AA compliant with proper contrast ratios
Responsive: Works beautifully on desktop, tablet, and mobile

See design_guidelines.md for complete design specifications.

🔧 Advanced Features

Voice Input

Powered by the Web Speech API for hands-free interaction:

Click the microphone icon in the chat input
Speak your health question naturally
Real-time speech-to-text transcription
Visual feedback while recording

Export Chat History

Save your conversations for future reference:

PDF: Professional formatting with timestamps
Text: Plain text format for easy reading
JSON: Machine-readable format for data analysis

Bookmarks

Mark important messages for quick access:

Click the star icon on any message
Bookmarks persist using localStorage
Yellow star indicates saved messages

Quick Actions

Pre-configured health topic cards for common queries:

💊 Medication Information
🩺 Symptoms Checker
❤️ Heart Health
🧠 Mental Health Support
🥗 Nutrition Advice
🏥 Health Checkups

📁 Project Structure

neural-rx/
├── client/
│   └── src/
│       ├── components/        # React components
│       │   ├── ChatMessage.tsx
│       │   ├── ExportDialog.tsx
│       │   ├── QuickActions.tsx
│       │   └── ScrollToBottomButton.tsx
│       ├── contexts/          # React contexts
│       │   └── BookmarksContext.tsx
│       ├── hooks/             # Custom hooks
│       │   ├── useVoiceRecognition.ts
│       │   └── useBookmarks.ts
│       ├── lib/               # Utilities
│       │   ├── exportUtils.ts
│       │   └── queryClient.ts
│       ├── pages/             # Page components
│       │   └── Home.tsx
│       └── types/             # TypeScript types
│           └── speech.d.ts
├── server/
│   ├── data/                  # Healthcare dataset
│   │   └── healthcare-conditions.ts
│   ├── elastic.ts             # Elasticsearch client
│   ├── gemini.ts              # Gemini AI integration
│   ├── routes.ts              # API endpoints
│   ├── storage.ts             # In-memory storage
│   └── index.ts               # Server entry point
├── shared/
│   └── schema.ts              # Shared TypeScript types
├── design_guidelines.md       # Design system documentation
├── replit.md                  # Project documentation
└── README.md                  # You are here!

🔌 API Endpoints

Chat & Search

POST /api/chat - Send message, get AI response + search results
GET /api/search?q={query} - Direct search without AI
GET /api/chat/history/:sessionId - Retrieve chat history

Analytics & Health

GET /api/analytics/searches - Search analytics data
GET /api/health - Server and Elasticsearch status

🧪 Testing

The application includes comprehensive end-to-end tests covering:

Chat functionality with AI responses
Search results rendering
Voice input interactions
Bookmark persistence
Export functionality
Dark mode toggling
Responsive design

Run tests:

npm test

🤝 Contributing

Contributions are welcome! Please follow these steps:

Fork the repository
Create a feature branch (git checkout -b feature/amazing-feature)
Commit your changes (git commit -m 'Add amazing feature')
Push to the branch (git push origin feature/amazing-feature)
Open a Pull Request

Please ensure your code follows the existing style and includes appropriate tests.

📝 License

This project is licensed under the MIT License - see the LICENSE file for details.

🙏 Acknowledgments

Elastic for providing powerful hybrid search capabilities
Google Cloud for Gemini AI's natural language understanding
Replit for the development environment and hosting platform
Shadcn UI for beautiful, accessible component primitives
The open-source community for amazing tools and libraries

📧 Contact

Project Maintainer: [Your Name]
Email: your.email@example.com
Project Link: https://github.com/yourusername/neural-rx

🎯 Hackathon Submission

This project was built for the Elastic + Google Cloud Hackathon Challenge.

Challenge: Build the future of AI-powered search using Elastic's hybrid search and Google Cloud's generative AI tools.

What makes Neural RX special:

✅ True hybrid search (keyword + semantic)
✅ Context-aware AI conversations
✅ Production-ready features (voice, export, bookmarks)
✅ Beautiful, accessible healthcare UI
✅ Real medical database with structured data
✅ Sub-8-second response times

Built with ❤️ for better healthcare information access

⭐ Star this repo if you find it helpful!

Report Bug · Request Feature · Live Demo

Name		Name	Last commit message	Last commit date
Latest commit History 6 Commits
CloudElasticAI		CloudElasticAI
LICENSE		LICENSE
README.md		README.md

License

Ibrahim4594/Neural-RX

Folders and files

Latest commit

History

Repository files navigation