InfoRAG

---

InfoRAG is an AI-powered documentation assistant that leverages Retrieval-Augmented Generation (RAG) to provide accurate, context-aware responses to professional queries across various domains.

Key Features:

• Knowledge-Based Ingestion Pipeline
• Vector Embedding
• Multimodal Input Processing
• Chain-of-Thought Implementation to reduce hallucinations

Overview:

Tech Stack:

Languages:

• Backend Framework:
  • FastAPI
• Frontend:
  • NextJS
  • TypeScript

Models:

• Gemini
• Deepseek R1 Distill Llama 70B
• Nomic-Embed-Text

Libraries:

• LlamaIndex:
• Gemini Genai SDK:
• Groq
• AstraDB

LlamaIndex Plugins:

• Tool Calling
• Arxiv Plugin
• PubMed Document API

Toolchain:

• uv Package Manager
• Docker

Usage:

1. Choose Domain
2. Enter your Prompt
3. Enable Web Search(Stock Prices, Weather, API Requests, Current News)
4. Audio Input
5. Attach PDF/Image

docker run -d -p 8000:8000 -v $(pwd)/.env:/app/.env sasquatch06/rag-backend:latest