Value AI

This app (value AI) allows users to upload an image of a vehicle and find out an estimated market value of that car. Value AI turns anyones camera into a car-valuation tool. It is a full-stack web app based on Python and React+Vite using FastAPI and Open AI GPT-4o vision API to identify make, model, and year to return up to the minute USD estimates.

Value AI Full-Stack Setup

*THE FOLLOWING INSTRUCTIONS ARE TEMPORARY AND ARE FOR ANYONE WHO WANTS TO TRY OUT THE SYSTEM BEFORE I HAVE IT HOSTED ONLINE. YOU CAN ALSO WATCH THE DEMO VIDEO ABOVE TO SEE IT IN ACTION. IT WILL BE AVALIABLE ONLINE SOON!*

This document summarizes the minimal configuration and code snippets for the React frontend and FastAPI backend so that the VITE_API_URL env var, fetch call, and JSON keys all align correctly. Feel free to reach out with any questions.

Quick Start: Testing Value AI Locally

Follow these steps to clone the repo, configure your environment, spin up both servers, and try the full upload→estimate flow yourself.

1. Clone & Enter

git clone https://github.com/tylermcclelland/valueai.git
          cd valueai

2. Configure Secrets

Create a file named .env in the project root (next to backend/ and frontend/):

OPENAI_API_KEY=sk-<your-key-here>
          VITE_API_URL=http://localhost:8000

(You would need to sign up at platform.openai.com to generate your own key.)

3. Start the Backend

Create & activate a Python virtual environment:
python3 -m venv .venv source .venv/bin/activate
Install dependencies:
pip install fastapi uvicorn python-dotenv openai
Run the FastAPI server on port 8000:
uvicorn backend.main:app --reload --port 8000

4. Start the Frontend

Install Node.js (if not already installed):
macOS: brew install node
Enter the frontend folder and install:
cd frontend npm install
Run the Vite dev server on port 5173:
npm run dev

5. Test via Browser

Open http://localhost:5173 in your browser.
Click “Upload Image”, select a car photo, then click “Get Value.”
The estimated USD value should appear below the buttons.

6. Test via cURL (optional)

curl -X POST http://localhost:8000/upload \
            -F 'image=@/full/path/to/car.jpg'

A JSON response like {"value":"2019 Toyota Camry — approx. $17,000"} indicates success.

If you run into CORS or missing-key errors, verify:

Your .env is in the root and matches the variable names exactly.
Both servers are running in separate terminals with the virtualenv active for the backend.
The browser console network tab for any failed requests.

Car Market-Value Estimator