👁️ Drishti Access – Event Situations Simulation Chatbot

Drishti Access is a multilingual, AI-powered chatbot designed to simulate and respond to crowd management scenarios during large public events. Using Google's Gemini API and Text-to-Speech, it lets you define venue structure, crowd size, gate timings, and ask realistic "what-if" questions like:

“What happens if Gate A is closed at 12 PM while Gate B remains open?”

“Imagine 2000 people came more. How should we make them disperse properly at 2PM to avoid stampede?”

It also reads out responses in the selected Indian language!

---

🚀 Features

• 🧠 Gemini-Powered Simulation
  Responds intelligently to what-if crowd management scenarios.

• 🗺️ Event Setup Panel
  Define venue name, structure, gate timings, timelines, and crowd size.

• 🗣️ Multilingual Support
  Supports 10 Indian languages for both response and TTS playback:

  • Hindi, Bengali, Marathi, Tamil, Telugu, Gujarati, Kannada, Malayalam, Punjabi, English

• 🔊 In-Browser TTS
  Uses gTTS to play voice feedback for every response.

• 🧾 Custom Gemini API Key Option
  Use your own API key securely via the sidebar.

---

💡 Topics Covered

• Streamlit UI Development
• Google Gemini (Generative AI) Integration
• Event Simulation Logic
• Text-to-Speech (gTTS)
• Multilingual Natural Language Processing
• In-Memory Audio Generation (io.BytesIO)

---

🛠️ Installation

1. Clone the repository:

   git clone https://github.com/your-username/drishti-access-chatbot.git
   cd drishti-access-chatbot
   

2. Install dependencies:

   pip install -r requirements.txt
   

3. Set up secrets: Create a .streamlit/secrets.toml file and add your Gemini API key:

   GEMINI_API_KEY = "your-api-key-here"
   

4. Run the app:

   streamlit run chat.py
   

---

📺 Demo

---

Made by Shivam Gupta for Google Agentic AI Day Hackathon