🧠 CentralGPT Backend

One Nexus Point to access and chat with multiple LLM providers — Mistral, Ollama, Groq, TogetherAI — with GitHub & PDF RAG capabilities.

Supports memory, chat history, persistent messages, and file retrieval from GitHub.

---

🛠 Tech Stack

Layer	Tech
Framework	FastAPI
LLM Framework	LlamaIndex
Validation	Pydantic
Database	Redis, PostgreSQL, Qdrant
Containerization	Docker
Deployment	Azure App Service
Git Integration	PyGithub

---

🚀 Features

• 🔁 Chat with multiple LLM providers
• 💾 Persistent chat history (PostgreSQL, Redis)
• 🧠 Session memory support
• 🔄 Dynamic model switching (Ollama, Groq, Mistral, etc.)
• 📂 GitHub Integration: Fetch files from any public/private repository using username , repo name,branch and commit
• 📄 PDF & Document RAG: Upload PDFs and retrieve context-aware answers
• 🔍 Vector search via Qdrant for fast retrieval

---

🧩 Setup Instructions

1. Clone the Repository

git clone https://github.com/DineshThumma9/centralGPT-backend.git
cd centralGPT-backend

2. Install Dependencies

pip install -r requirements.txt

3. Configure Environment Variables

Create a .env file in the root directory:

DATABASE_URL=
REDIS_URL=
QDRANT_URL=
QDRANT_API_KEY=
GROQ_API_KEY=
COHERE_API_KEY=
HUGGINGFACE_HUB_TOKEN=
GITHUB_TOKEN=

4. Frontend Setup (Required)

Clone and set up the frontend:

git clone https://github.com/DineshThumma9/centralGPT.git

Follow the frontend repo instructions to configure and run it.

---

🧪 Development (Local)

uvicorn src.main:app --reload 

---