Documind

Documind is a SaaS application that combines document management with conversational AI, powered by generative AI models like GPT-3.5. It enables organizations to create searchable knowledge bases from their documentation and allows users to interact with documents using natural language queries.

Table of Contents

• Features
• How Documind Works
• Getting Started
• Configuration
• Project Structure
• Usage Example
• Deployment
• Acknowledgments
• License
• Contributing

---

Features

• Document crawling and ingestion
• Vector embedding creation and storage (Pinecone)
• Conversational search powered by generative AI (GPT-3.5 and others)
• Real-time chat interface
• PDF viewing and file upload
• Subscription management (Stripe integration)
• Secure authentication (sign-in/sign-up)
• Modular and scalable Next.js architecture

---

How Documind Works

1. Crawling Documentation Website
   • Documind crawls the provided documentation URLs to collect content.
2. Creating Knowledge Base
   • Extracted content is converted into vector embeddings.
   • Embeddings are stored in Pinecone for fast similarity search.
3. Search Process
   • User queries are embedded and compared against the knowledge base.
   • Relevant context is retrieved and passed to generative AI models (e.g., GPT-3.5).
   • The AI generates conversational answers based on the documentation.

---

Getting Started

Prerequisites

• Node.js (v18+ recommended)
• npm, yarn, or pnpm
• Pinecone account (for vector database)
• OpenAI API key (for generative AI)
• Stripe account (for subscriptions, if needed)

Installation

1. Clone the repository:

   git clone https://github.com/your-repo/documind.git
   cd documind

2. Install dependencies:

   npm install
   # or
   yarn install

3. Create a .env.local file and add your API keys:

   PINECONE_API_KEY=your-pinecone-api-key
   PINECONE_INDEX_NAME=your-pinecone-index
   PINECONE_ENVIRONMENT=your-pinecone-environment
   OPENAI_API_KEY=your-openai-api-key
   STRIPE_SECRET_KEY=your-stripe-secret-key
   

4. Start the development server:

   npm run dev
   # or
   yarn dev

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

---

Configuration

To create a knowledge base, provide:

• Documentation Website URL
  Example: https://nextjs.org/docs
• Documentation Website URL Match
  Example: https://nextjs.org/docs/**
  Use ** as a wildcard for URL patterns.
• CSS Selector for Main Text Content
  Example: .main-content
  Helps extract relevant content from documentation pages.
• Pinecone Details
  • API Key
  • Index Name
  • Environment
• OpenAI API Key
  Used for generating conversational responses.

---

Project Structure

components.json
public/
	file.svg
	globe.svg
	next.svg
	vercel.svg
	window.svg
src/
	middleware.ts
	app/
		favicon.ico
		globals.css
		layout.tsx
		page.tsx
		api/
			chat/
				route.ts
			create-chat/
				route.ts
			get-messages/
				route.ts
			stripe/
				route.ts
			webhook/
				route.ts
		chat/
			[chatId]/
				page.tsx
		sign-in/
			[[...sign-in]]/
				page.tsx
		sign-up/
			[[...sign-up]]/
				page.tsx
		success/
			page.tsx
	components/
		ChatComponent.tsx
		ChatSidebar.tsx
		FileUpload.tsx
		MessagesList.tsx
		PDFViewer.tsx
		Providers.tsx
		SubscriptionButton.tsx
		ui/
			button.tsx
			input.tsx
	lib/
		context.ts
		embeddings.ts
		pinecone.ts
		s3-server.ts
		s3.ts
		stripe.ts
		subscription.ts
		utils.ts
		db/
			index.ts
			schema.ts

---

Usage Example

import { Documind } from "documind";

const documind = new Documind({
  documind_key: "your-documind-key",
});

const { answer, message, error } = await documind.search(searchQuery);

---

Deployment

Documind is designed for deployment on Vercel or similar platforms.

1. Push your code to a GitHub repository.
2. Connect your repository to Vercel.
3. Set environment variables in Vercel dashboard.
4. Deploy and access your app online.

---

Acknowledgments

Documind draws inspiration from BuilderIO/gpt-crawler, which focuses on crawling documentation websites to generate knowledge files for OpenAI assistants. Documind extends this by integrating conversational search directly into documentation sites using generative AI.

---

License

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

---

Contributing

Contributions are welcome! Please open an issue or submit a pull request for improvements or new features.