Vehicle Safety Complaint Chatbot

An AI-powered chatbot application for collecting vehicle safety complaints and feedback. Built with Streamlit and integrated with Hugging Face LLM APIs.

Features

• Intelligent Data Collection: Uses LLM to extract and validate vehicle safety information from natural conversation
• RAG (Retrieval Augmented Generation): Context-aware responses with domain knowledge about vehicles, VINs, and safety components
• Smart Field Management: Prevents duplicate questions and tracks collected information
• Data Validation: Real-time validation of VINs, state codes, dates, and other fields
• Google Sheets Integration: Automatic storage of complaints and feedback

Project Structure

├── app.py                 # Main Streamlit application entry point
├── complaint_bot.py       # Safety complaint collection chatbot
├── feedback_bot.py        # General feedback collection chatbot
├── shared_utils.py        # Shared utilities, LLM functions, and field management
├── knowledge_base.py      # RAG knowledge base for vehicle/safety context
├── requirements.txt       # Python dependencies
├── .streamlit/
│   └── secrets.toml       # API keys and credentials (not in repo)
└── README.md

Prerequisites

• Python 3.8 or higher
• Hugging Face API key
• Google Cloud service account (for Sheets integration)

Installation

1. Clone the repository:

   git clone https://github.com/Sharkyii/final-complaint-chatbot.git
   cd final-complaint-chatbot

2. Create a virtual environment:

   python -m venv venv
   source venv/bin/activate  # On Windows: venv\Scripts\activate

3. Install dependencies:

   pip install -r requirements.txt

4. Configure secrets: Create .streamlit/secrets.toml with:

   [huggingface]
   api_key = "your-huggingface-api-key"
   
   [gcp_service_account]
   type = "service_account"
   project_id = "your-project-id"
   private_key_id = "your-key-id"
   private_key = "-----BEGIN PRIVATE KEY-----\n...\n-----END PRIVATE KEY-----\n"
   client_email = "your-service-account@project.iam.gserviceaccount.com"
   client_id = "your-client-id"
   auth_uri = "https://accounts.google.com/o/oauth2/auth"
   token_uri = "https://oauth2.googleapis.com/token"

Usage

Run the application:

streamlit run app.py

The application will be available at http://localhost:8501

Configuration

Environment Variables

All sensitive configuration is managed through Streamlit secrets (.streamlit/secrets.toml):

Secret	Description
huggingface.api_key	Hugging Face API key for LLM access
gcp_service_account	Google Cloud service account credentials

Google Sheets Setup

1. Create a Google Sheet named "Safety_Reports"
2. Share the sheet with your service account email
3. The application will automatically append rows with collected data

API Reference

LLM Model

• Chat Model: meta-llama/Llama-3.1-8B-Instruct (via Hugging Face Router)

Data Fields

Complaint Fields:

• Vehicle: Make, Model, Model_Year, VIN, Mileage
• Location: City, State
• Incident: Speed, Crash, Fire, Injured, Deaths, Component
• Details: Description, Date_Complaint

Automated Fields:

• Timestamp, Input_Length, Suspicion_Score, User_Risk_Level

Development

Code Style

• Follow PEP 8 guidelines
• Use type hints where applicable
• Document functions with docstrings

Testing

python -m py_compile app.py complaint_bot.py shared_utils.py knowledge_base.py

License

This project is proprietary software. All rights reserved.

Support

For issues or questions, please open an issue in the repository.