Hospital AI Assistant

An AI-powered voice-based hospital receptionist built using FastAPI and Google Gemini. This system allows users to interact using voice, check doctor availability, and book appointments seamlessly.

---

Features

• Voice interaction using browser speech recognition
• AI-powered responses using Gemini
• Doctor listing and appointment booking
• Text-to-Speech responses
• Multi-language support (English, Hindi, Telugu)
• SQLite database for storing appointments

---

Tech Stack

• Backend: FastAPI (Python)
• Frontend: HTML, TailwindCSS, JavaScript
• AI Model: Google Gemini
• Database: SQLite
• Text-to-Speech: gTTS

---

Project Structure

hospital-ai/
│
├── app.py              # FastAPI backend
├── frontend.html       # UI for voice interaction
├── .env                # API keys (not pushed to GitHub)
├── README.md           # Project documentation

---

Setup Instructions

1. Clone the Repository

git clone <your-repo-url>
cd hospital-ai

---

2. Create Virtual Environment

python -m venv venv
.\venv\Scripts\activate

---

3. Install Dependencies

pip install fastapi uvicorn python-dotenv google-genai gtts

---

4. Setup Environment Variables

Create a .env file in the root directory:

GEMINI_API_KEY=your_api_key_here

---

5. Run the Server

uvicorn app:app --reload

---

6. Open in Browser

http://127.0.0.1:8000

---

Usage

1. Enter your name and email
2. Select preferred language
3. Start voice interaction
4. Ask about doctors or book an appointment

---

Future Improvements

• Deploy the application online
• Integrate real hospital APIs
• Add chat-based fallback UI
• Improve mobile responsiveness

---

👩‍💻 Author

Sanjana Capstone Project – Hospital AI Assistant