app.py

import gradio as gr
import google.generativeai as genai
import base64
import io
from PIL import Image
import os
import json

# Configure Google Cloud credentials (replace with your actual API key or setup)
genai.configure(api_key=os.getenv("GOOGLE_API_KEY"))

# Select the Gemini Pro Vision model
model = genai.GenerativeModel('gemini-pro-vision')

# Prompt definition
prompt = """
You are analyzing a medical document or an application form from patient.
Extract the following fields as JSON:
- Position applied for
- Office/Ministry
- Duty station
- First name(s)
- Surname
- Date of birth
- Gender
- Citizenship
- Postal Address
- Residential Address
- Email
- Phone number (mobile)
"""

def process_image(image: Image.Image):
    buffered = io.BytesIO()
    image.save(buffered, format="JPEG")
    base64_image = base64.b64encode(buffered.getvalue()).decode()

    contents = [
        genai.Content.create(role="user", parts=[genai.Part.from_text(prompt)]),
        genai.Content.create(role="user", parts=[genai.Part.from_data(base64.b64decode(base64_image), mime_type="image/jpeg")])
    ]
    response = model.generate_content(contents)
    return response.text

# Gradio interface
demo = gr.Interface(
    fn=process_image,
    inputs=gr.Image(type="pil"),
    outputs="textbox",
    title="Healthelic Form Data Extractor (Doc Scaner)",
    description="Upload a scanned medical form to extract key fields."
)

if __name__ == "__main__":
    demo.launch()