SwiftHerb - AI Pharmacist Assistant

A low-latency AI pharmacist assistant that converts chat queries into high-value iHerb commissions.

Tech Stack

• Frontend: Next.js 15 (App Router), Tailwind CSS
• Backend Infrastructure: Cloudflare Workers (Edge Computing)
• Data Persistence: Cloudflare Workers KV (Key-Value Store)
• AI Model: DeepSeek-V3 (via OpenRouter)
• Affiliate Integration: Impact.com (iHerb program; direct iHerb links until tracking is live)

Getting Started

1. Install dependencies:

pnpm install

2. Copy the environment variables:

cp .env.example .env

3. Fill in your environment variables in .env:

   • DEEPSEEK_API_KEY: Your DeepSeek API key
   • IMPACT_ACCOUNT_SID and IMPACT_AUTH_TOKEN: Impact.com API (server-side / scripts only)
   • NEXT_PUBLIC_SITE_URL: Your site URL (for development: http://localhost:3000)

4. Run the development server:

pnpm dev

5. Open http://localhost:3000 in your browser.

Project Structure

swiftherb/
├── app/                    # Next.js App Router
│   ├── api/               # API routes
│   ├── layout.tsx         # Root layout
│   ├── page.tsx           # Home page
│   └── globals.css        # Global styles
├── components/            # React components
│   ├── Header.tsx         # Sticky header with logo
│   ├── ChatInterface.tsx # Main chat UI
│   ├── MessageBubble.tsx  # Message display component
│   ├── ChatInput.tsx      # Chat input component
│   ├── ProductCard.tsx    # Product display card
│   └── DisclaimerBar.tsx  # Fixed disclaimer bar
├── lib/                   # Utility functions
│   ├── ai.ts             # AI integration (OpenRouter)
│   ├── kv.ts             # Cloudflare KV operations
│   ├── admitad.ts        # Admitad link generation
│   └── affiliate.ts     # Unified affiliate link generator
└── SWIFTHERB_MASTER_PLAN.md # Project master plan

Features

• AI-Powered Chat: Conversational interface powered by DeepSeek-V3
• Product Recommendations: AI suggests products wrapped in [[Double Brackets]]
• Affiliate Links: Automatic Admitad deep linking for all product recommendations
• Safety Features: Emergency keyword detection for medical emergencies
• Wellness Stacks: AI suggests multiple products for higher cart values

Next Steps

1. Set up Cloudflare Workers and KV namespace
2. Populate KV store with product data
3. Implement product metadata fetching from KV
4. Add error handling and loading states
5. Deploy to Cloudflare Pages

License

Private project - All rights reserved