🧠 Ethora RAG Demos

Retrieval-Augmented Generation (RAG) demos built with Ethora, LangChain, and OpenAI.
Quickly learn how to build knowledge-based AI assistants that can index your own website or documents.

Part of the Ethora SDK ecosystem — see all SDKs, tools, and sample apps. Follow cross-SDK updates in the Release Notes.

---

🚀 Overview

This repository contains practical examples showing how to connect Ethora’s AI chat widget and backend tools with RAG pipelines.
Each demo is designed to help developers understand how to:

• Crawl and index web pages or documents
• Store embeddings in a vector database
• Retrieve relevant chunks based on user queries
• Use LLMs (OpenAI, Anthropic, etc.) to generate contextual answers
• Serve the results via Ethora’s chat component or WordPress plugin

You can use these demos as templates for your own AI assistants, knowledge bases, or support chatbots.

---

📦 Demos Included

Demo	Description
basic_rag_openai	Minimal RAG pipeline using OpenAI embeddings and text completion.
rag_with_site_crawler	End-to-end demo combining the site_crawler tool for automated website indexing.
rag_langchain	Example using LangChain Retriever + Chain for custom prompt templates.
rag_api_server	Demonstrates how to serve results to your Ethora Chat Component or WordPress plugin.

---

🧰 Tech Stack

• Python 3.10+
• LangChain
• OpenAI API (or other LLM providers)
• FAISS / ChromaDB
• Ethora Chat Component (for front-end embedding)
• Ethora Site Crawler (for ingestion)

---

⚙️ Quick Start

Clone the repo and install dependencies:

git clone https://github.com/dappros/rag_demos.git
cd rag_demos
pip install -r requirements.txt

Set your API keys (for example, OpenAI):

export OPENAI_API_KEY="sk-..."

Run any demo:

python demos/basic_rag_openai.py

You can also index your website before running:

python ../site_crawler/crawl.py https://yourwebsite.com

Then query your knowledge base using the chat widget or CLI interface.

---

💬 Connect to Ethora Chat Widget

Once you have a RAG backend running, you can embed it into any website:

<script src="https://cdn.jsdelivr.net/npm/ethora-chat-component@latest/dist/widget.js"></script>
<script>
  EthoraChatWidget.init({
    endpoint: "https://your-server/api/rag",
    title: "Ask our AI Assistant",
  });
</script>

You can also use the WordPress plugin for no-code integration.

---

🌍 Related Repositories

Project	Description
ethora-chat-component	Embeddable AI chat widget built with React & TypeScript.
ethora-wp-plugin	WordPress plugin for adding AI chat assistants.
site_crawler	Lightweight crawler and indexer for website content.
ethora	Main Ethora platform – low-code engine for chat, AI, and Web3 apps.

---

🧭 Roadmap

• Add demo for multi-vector store (FAISS + Chroma)
• Add support for PDF & DOCX ingestion
• Add streaming chat UI example
• Integrate with local open-source LLMs (Mistral, Ollama, etc.)
• Add Docker-based RAG starter kit

---

🤝 Contributing

Contributions are welcome! To contribute:

1. Fork this repo
2. Create a new branch
3. Submit a pull request with a clear description

For larger features, please open an issue first to discuss your idea.

---

🧾 License

This project is open source under the MIT License (content of websites used in demos is not included and belongs to its original owners).

---

💡 About Ethora

Ethora is a low-code “super-app” engine for chat, AI assistants, digital wallets, and communities. It allows developers to rapidly build apps that combine messaging, AI, and Web3 in one stack.

• 🌐 ethora.com
• 💬 Community Forum
• 🧰 Developer Docs

---

If you find this repo useful, please ⭐ star it and share your RAG demo with us!

---

✅ To use:

1. Go to your repo → Add file → Create new file → README.md
2. Paste everything above
3. Commit to main
4. Optionally run git add README.md && git commit -m "add improved README" locally.

---