Skip to content

skushagra9/rag-pdf

Folders and files

NameName
Last commit message
Last commit date

Latest commit

Β 

History

23 Commits
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 

Repository files navigation

RAG System - Superteam Vietnam AI Assistant

🎯 Objective

Superteam Vietnam currently relies on manual effort to manage its communication channels like Telegram and Twitter. This project aims to develop an AI-driven solution to automate, enhance, and streamline content creation, management, and community interactions.

This MVP solution lays the foundation for a fully-fledged AI system.

βœ… How This RAG System Addresses the Requirements

πŸ“Œ 1. Telegram Knowledge Portal Bot

βœ” Solution Implemented:

  • Developed a Telegram bot that acts as a knowledge base for Superteam Vietnam.
  • Uses Retrieval-Augmented Generation (RAG) to ensure accurate responses.
  • Ensures no hallucinationsβ€”the bot confidently says "NO" if it does not find a relevant answer.
  • Integrated Admin UI for document uploads to continuously train and improve the bot.

πŸ“Œ 3. Twitter Management & Content Advisor Assistant

βœ” Solution Implemented:

  • The AI integrates with Superteam Vietnam’s Twitter account to:
    • Propose tweets for human approval.
    • Help refine drafts by:
      • Suggesting keywords.
      • Correcting Twitter handles based on Superteam VN’s followed accounts.
    • Finalize and publish tweets.
    • Uses local AI models for faster & more relevant suggestions.

πŸ“Œ 4. Local LLM Deployment

βœ” Solution Implemented:

  • Deployed & runs locally to ensure data privacy.
  • Uses Ollama to run LLM models on local machines.

Installation Steps:

# Install Ollama
curl -fsSL https://ollama.ai/install.sh | sh  

# Pull the Llama model
ollama pull llama3.2:1b

# Run the model
ollama run llama3.2:1b

πŸ”— Admin UI Installation Guide
πŸ”— Telegram Stack Installation Guide
πŸ”— Python Backend Installation Guide