📚 RAG Demo App (Ruby on Rails + Gemini/OpenAI)

A simple, clean Retrieval-Augmented Generation (RAG) app built in Ruby on Rails, using Gemini/OpenAI for embeddings and generative AI.

---

🚀 Features

• Upload .txt / .pdf / .docx / .html/ .rb/ .py/ .md documents as knowledge base
• Automatic document chunking & vector embeddings (Gemini API/ OpenAI API)
• Semantic search powered by pgvector
• LLM-generated answers grounded in uploaded documents
• Dashboard for content/documents management
• Clean UI with TailwindCSS + Slim

---

🧠 How It Works

1. Upload Documents
   Uploads are chunked and stored with vector embeddings.

2. Query the System
   Users ask natural-language questions.

3. Search & Retrieval
   The app uses semantic search to find relevant chunks.

4. Generate Answers
   Retrieved chunks are sent to Gemini or OpenAI to generate grounded, accurate responses.

---

⚙️ Tech Stack

• Ruby on Rails 8
• pgvector + PostgreSQL
• Gemini API (Google AI)/ OpenAI API
• TailwindCSS + Slim

---

📥 Setup Instructions

git clone https://github.com/himalayan-sanjeev/retrieval-augmented-generation-rails
cd retrieval-augmented-generation-rails

bundle install
yarn install

# Add your GEMINI_API_KEY or OPENAI_ACCESS_TOKEN to .env

rails db:create db:migrate
rails server