🐹 Go-RAG: Extreme AI Intelligence Engine

---

⚡ The Project Overview

This is a High-Performance RAG (Retrieval-Augmented Generation) system built in Golang. It doesn't just chat; it thinks by connecting your PDFs and live URLs to a massive vector brain.

🛡️ Why This Stack?

• Golang: For lightning-fast concurrent processing of large PDFs.
• Astra DB (DataStax): Serverless Vector DB for global-scale AI retrieval.
• Supabase/PG: The backbone for user accounts, session history, and relational data.
• Groq: Sub-second LLM inference (Llama-3/Mixtral).
• Voyage AI: State-of-the-art text embeddings.

---

🚀 Core Features

Feature	Description	Status
🔐 Full Auth	Google OAuth2 + JWT Integration	✅ Done
📄 PDF Ingestion	Extract, Chunk, and Embed complex PDF documents	✅ Done
🌐 URL Scraper	Live web scraping for real-time AI context	✅ Done
🧠 Vector Search	Semantic retrieval using Astra DB & Voyage AI	✅ Done
💬 Chat History	Persistent sessions stored in Supabase/PostgreSQL	✅ Done
👮 Admin Panel	Full control over users, data, and embedding indices	✅ Done

---

🛠️ System Architecture

1. Ingestion: Files/Links are processed via Go Goroutines.
2. Vectorization: Text chunks $\rightarrow$ Voyage AI $\rightarrow$ High-dimensional Vectors.
3. Storage: Vectors live in Astra DB; Meta-data lives in Supabase.
4. Retrieval: User Query $\rightarrow$ Vector Search $\rightarrow$ Context Injection $\rightarrow$ Groq Answer.

---

📦 Environment Setup

Create a .env file in your root folder:

# Database & Auth
DATABASE_URL=postgresql:supabase_url
JWT_KEY=any_secret_key
PORT=8080
APP_ENV=production

# AI Keys (Astra & Inference)
ASTRA=your_astra_token
ASTRA_END_POINT=your_endpoint_here
VOYAGE_API_KEY=your_voyage_key
GROQ_API_KEY=your_groq_key

# Integrations
API_BERVO=your_brevo_key
EMAIL=your_admin_email
PASSWORD=your_app_password
GOOGLE_CLIENT_ID=your_google_id

🚦 Getting Started

1. Installation

go mod download

2. Run the Engine

go run ./cmd/main.go