Antara AI - A Memory-Driven Conversational AI

Antara AI is a sophisticated conversational AI system built on the LangGraph framework, featuring persistent memory, student-centric features, gamification, and advanced crisis detection. This project aims to provide a supportive and engaging conversational experience for students, helping them manage academic stress and peer pressure.

Features

• Memory-Driven Conversations: The agent remembers past interactions and learns from them, providing a personalized experience.
• Student-Centric Features: Includes a RAG pipeline to provide information on study materials and academic coping strategies.
• Gamification: A "Snap Streak" and XP system to encourage user engagement.
• Advanced Crisis Detection: A multi-layered sentiment analysis system to detect and respond to user distress.
• Modular and Extensible: Built with a clean architecture that allows for easy extension and integration.

Architecture Overview

The application is built with a modular, layered architecture:

• UI Layer (streamlit_app.py, cli_app.py): User interface components for interacting with the agent.
• API Layer (api.py): A FastAPI-based REST API that exposes the agent's functionality.
• Service Layer (core/service.py): A high-level interface that orchestrates the core logic.
• Core Layer (core/): Contains the core agent logic, including the LangGraph state machine, memory management, and tools.
• Configuration Layer (config/): Centralized configuration for the application, including prompt templates and system settings.

Setup and Installation

Prerequisites

• Python 3.10+
• MongoDB
• An API key for an embedding provider (e.g., OpenAI, Google)

Installation

1. Clone the repository:

   git clone https://github.com/Ishwarpatra/AntaraAI.git
   cd AntaraAI

2. Create and activate a virtual environment (recommended):

   python3 -m venv venv
   source venv/bin/activate

3. Install dependencies: The project is missing a requirements.txt file. You will need to manually install the following dependencies:

   pip install fastapi uvicorn pymongo langchain langgraph langchain-openai langchain-google-genai langchain-community sentence-transformers python-dotenv bandit tenacity pytest

   Note: This list is based on the issues encountered during development. It might not be exhaustive.

4. Set up environment variables: Create a .env file in the root of the project and add the following:

   # MongoDB
   MONGODB_URI="mongodb://localhost:27017/"
   DATABASE_NAME="antara_ai"
   
   # Embedding Provider (choose one)
   EMBEDDING_PROVIDER="openai" # or "google" or "huggingface"
   OPENAI_API_KEY="your-openai-api-key"
   # GOOGLE_API_KEY="your-google-api-key"
   
   # Crisis Contact (for crisis_escalation_tool)
   EMERGENCY_CONTACT_PHONE="+1234567890" # Example phone number for Twilio
   

5. Ingest academic resources: Run the ingestion script to populate the academic resources knowledge base:

   python utils/ingestion.py

Running the Application

• API Server:

  uvicorn api:app --host 127.0.0.1 --port 8000 --reload

• Streamlit UI:

  streamlit run streamlit_app.py

• CLI App:

  python cli_app.py

API Reference

The API is documented using OpenAPI (Swagger). Once the API server is running, you can access the interactive documentation at http://127.0.0.1:8000/docs.

Key Endpoints

• POST /chat: Process a user message and get a response.
• POST /mood/log: Log the user's mood and get a gamification update.
• GET /users: Get a list of users.
• POST /users: Create a new user.

Usage Examples

CLI

python cli_app.py

API (using curl)

• Create a user:

  curl -X POST "http://127.0.0.1:8000/users" -H "Content-Type: application/json" -d '{"user_id": "test_user"}'

• Send a message:

  curl -X POST "http://127.0.0.1:8000/chat" -H "Content-Type: application/json" -d '{"user_id": "test_user", "thread_id": "some_thread", "message": "Hello, how are you?"}'

Contribution Guidelines

Please see CONTRIBUTING.md for details on how to contribute to this project.