CDS Policy Vision

A Flask web application that analyzes parking sign images and provides parking guidance using Google's Gemini AI.

What It Does

1. Upload a parking sign image
2. Enter vehicle details (type, date, time)
3. Get two AI-powered analyses:
   • CDS (Curb Data Specification) output - structured parking rule data
   • Plain English parking guidance - "Can I park here?" answer

Quick Start

Local Development

# Install dependencies
pip install flask google-generativeai python-dotenv

# Set up environment
echo "API_KEY=your_gemini_api_key" > .env

# Run the app
python3 app.py

Visit http://localhost:5000

Docker

# Build
docker build -t my-flask-app .

# Run
docker run -p 8080:8080 --env-file .env my-flask-app

Visit http://localhost:8080

Required Files

• system_instructions1.txt - Instructions for CDS analysis
• system_instructions2.txt - Instructions for parking guidance
• prompt1.txt - Prompt for first API call
• prompt2.txt - Template for second API call (uses user variables)
• .env - Your Gemini API key

How It Works

1. First API Call: Analyzes sign image → generates structured CDS data
2. Second API Call: Uses sign image + user input → provides parking advice

Built with Flask, Google Gemini AI, and Tailwind CSS.