FlySafe.ai ✈️

FlySafe.ai is a personal AI-powered inspection agent that uses Amazon Nova Pro and Amazon Nova Sonic to assist with video analysis, natural language interaction, and tool-based automation.

Here's the video demo.

Features ✨

Video Analysis (Nova Pro)

• Uses the Amazon Nova Pro model to analyze inspection videos.
• Automatically generates a detailed inspection report.

Conversational Agent (Nova Sonic)

• Uses the Amazon Nova Sonic model to:
  • Answer user questions about the inspection.
  • Perform tool-based actions using integrated APIs.

Tool-Calling Functionalities 🔧

1. Google Shopping Search

• Triggered when the user asks the agent to look for any replacement parts or products.
• Uses the SERP API to fetch and display the top 3 results from Google Shopping.

2. Invoice Generation

• Based on the selected product from the search results.
• Automatically generates a PDF invoice with the relevant info

3. Email Sending

• Uses the SendGrid API.
• Sends the generated invoice PDF and the conversation transcript to the user’s email on confirmation

Usage 📦

Before you begin, make sure you have the following:

• Python 3.13+
• API keys: (check .env file for more info)

Now follow these steps to use the project-

1. Clone the repository

   git clone https://github.com/JayShukla8/FlySafe.ai.git
   cd FlySafe.ai
   

2. Create a virtual env and activate it

    python -m venv .venv
   
    # Activate the virtual environment
   
    # On Windows:
    .venv\Scripts\activate
   
    # On macOS/Linux:
    # source .venv/bin/
   

3. With the virtual environment activated, install the required packages:

   python -m pip install -r requirements.txt --force-reinstall
   

4. Configure AWS credentials:

   The application uses environment variables for AWS authentication. Set these before running the application:

   export AWS_ACCESS_KEY_ID="your-access-key"
   export AWS_SECRET_ACCESS_KEY="your-secret-key"
   export AWS_DEFAULT_REGION="us-east-1"
   

5. Specify the user Email

   Please specify the email which will send the notifications at line.

6. Start

   Start the interface via the command

   streamlit run src/workflow/interface_streamlit.py
   

Contribution 🌱

This project will remain open for contribution, please feel free to help.

Acknowledgments 🤜🤛

Thanks to Pavan, Shubham and Manan for helping throughtout.

I have made changes to this code for running the Nova Sonic model. Refer to its Readme file for more details.