/LLAMA_Hackathon

An Interview Prep Advisor

Primary LanguageTypeScript

LLAMA Hackathon Project

This project provides an end-to-end interactive question-and-answer experience. It includes:

  • Frontend (UI) for user interaction and silence detection.
  • Backend 1 (server.py) for scraping and question management.
  • Backend 2 (app.py) for two-way communication, real-time audio feedback, and Groq and Llama integration.

How to Run the Project

Frontend (UI)

Overview

The frontend handles user interactions, silence detection, and communicating with the backend.

Features

  1. Silence Detection:
    • The frontend monitors user input activity during sessions.
    • If no input is detected within a defined threshold, it automatically requests the next question from the backend.
  2. Interactive Display:
    • Dynamically renders questions and processes user answers in real-time.

Setup Instructions

  1. Navigate to the ui directory: 
    cd ui
  2. Install dependencies: 
    npm install
  3. Start the development server: 
    npm run dev
  4. Open the provided URL (e.g., http://localhost:3000) in your browser to access the UI.

Backend 1: server.py

Overview

  • Manages scraping and question updates.
  • Built with FastAPI and Selenium for browser automation.

Setup Instructions

  1. Navigate to the Backend Directory:

    cd src/backend

  2. Install Required Dependencies:

    pip install fastapi pydantic selenium python-dotenv

  3. Ensure WebDriver is Installed:

    • Install a browser WebDriver (e.g., ChromeDriver for Chrome).
      • For Ubuntu: 
        sudo apt install chromedriver
      • Download manually from here.

  4. Set Environment Variables:

    • Add a .env file in the backend directory with API keys or configurations (if required): 
      API_KEY=your_api_key_here

  5. Run the Backend:

    python server.py

Backend 2: app.py

Overview

  • Acts as a middleware for real-time communication between the frontend and backend.
  • Facilitates:
    • User actions like submitting answers and fetching the next question.
    • Real-Time Audio Connection: Integrates Groq and Llama to provide live audio-based interactions and feedback for mock interviews and conversational scenarios.

Features

  1. Groq and Llama Integration:

    • Connects to Groq and Llama APIs for generating real-time responses for user inputs.
    • Provides intelligent feedback and guidance, simulating interviews or conversational scenarios.

  2. Real-Time Audio Feedback:

    • Uses ElevenLabs or AssemblyAI for text-to-speech and speech-to-text capabilities.
    • Streams responses back to the user in real-time, enabling interactive audio sessions.

  3. User Interaction Flow:

    • Frontend sends user queries or audio to app.py.
    • app.py processes the input with Groq and Llama, sending back text or audio feedback.

Setup Instructions

  1. Navigate to the Backend Directory:

    cd src/backend

  2. Install Required Dependencies:

    pip install fastapi requests python-dotenv elevenlabs assemblyai groq

  3. Set Environment Variables:

    • Add a .env file in the backend directory: 
      XI_API_KEY=your_elevenlabs_api_key_here
GROQ_API_KEY=your_groq_api_key_here
AAI_API_KEY=your_assemblyai_api_key_here
LLAMA_API_KEY=your_llama_api_key_here

  4. Run the Middleware:

    python app.py

Project Workflow

1. Start the Frontend:

  • The UI dynamically renders questions and monitors user activity.
  • Silence detection triggers a transition to the next question if inactivity exceeds the threshold.

2. Backend 1 (server.py) Handles Core Logic:

  • Scrapes questions from an external source.
  • Maintains and updates the question bank based on user progress.

3. Backend 2 (app.py) Enables Advanced Features:

  • Relays user requests (e.g., audio input or questions) from the frontend to the backend.
  • Processes the input with Groq and Llama for intelligent responses.
  • Converts responses into real-time audio feedback for enhanced interactivity.

Future Improvements

  1. Frontend:

    • Improve the silence detection algorithm for better sensitivity.
    • Add real-time WebSocket support for instant updates.

  2. Backend:

    • Optimize scraping for reliability and speed.
    • Enhance Groq and Llama API interactions for more robust responses.

  3. End-to-End:

    • Dockerize the application for seamless deployment.
    • Add metrics to analyze user behavior and improve feedback quality.

This setup ensures a fully interactive Q&A and feedback system with advanced audio and real-time features powered by Groq and Llama.