Skip to content

ozcanmiraay/llama_hackathon

 
 

Repository files navigation

BabAI - AI-Powered WhatsApp Bot for Social Security Information

babAI-new-version

🚀 Project Overview

BabAI is a WhatsApp chatbot that provides users with accurate, up-to-date information on social security and pension regulations in Bulgaria. It leverages AI to retrieve information directly from official sources and ensure high-quality, reliable responses. Built with convenience, accuracy, and safety in mind, BabAI empowers users to access critical information anytime, anywhere.

🎯 Key Features

  • Seamless WhatsApp Integration: Available on a platform familiar to most Bulgarians, with a 90% adoption rate.

  • Real-Time Accurate Information: Fetches and presents information from National Social Security Institute documents.

  • Guardrails for Safety: Ensures responses are ethical, accurate, and compliant with legal guidelines.

🛠️Tech Stack

  • Backend: FastAPI (Webhooks), Uvicorn, Ngrok (Reverse Proxy), Twilio API (for WhatsApp integration)

  • Database: Pinecone Vector Database

  • AI Frameworks: LLama, Together AI, Hugging Face

  • DevOps: GitHub Actions for CI/CD, Heruko

  • Other Tools: Python (Poetry for dependency management)

🏗Project Architecture

  1. User Input: Users send questions about pensions or social security via WhatsApp
  2. API Gateway: Twilio API captures and forwards user queries to Ngrok reverse proxy
  3. Reverse Proxy: Ngrok forwards payload to BabAI's FastAPI backend
  4. Information Retrieval: Backend retrieves accurate, context-specific data from official documents.
  5. Response Generation: The AI model generates responses, filtered by guardrails for quality and safety.
  6. Output: The user receives a comprehensive, user-friendly answer directly on WhatsApp via webhook.

🚀 Quick Start

Prerequisites:

  • Python 3.9+ installed
  • Twilio Account for WhatsApp API integration
  • Ngrok
  • Together AI API

Installation

Clone repository

git clone https://github.com/username/babai.git
cd babai

Install dependencies:

poetry install

Set up environment variables:

cp .env.example .env

Populate .env with your credentials (see the Environment Variables section below).

Run the server locally:

uvicorn app.main:app --reload

Start ngrok reverse proxy:

ngrok http http://localhost:<port-number>

Start chatting with BabAI ✨✨

🙌Team (alphabetic order)

  • Gabriela Tsvetkova
  • Miray Özcan
  • Saad Asad
  • Tetiana Bass
  • Hackathon Mentor from Meta

🗣Acknowledgements

Thanks to Hack for Social Impact and all team members who helped make BabAI a reality! Thanks mom for giving us domain expertise on this complex topic 💗

About

No description, website, or topics provided.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • Python 99.8%
  • Procfile 0.2%