Product Pipeline

This project is an ai based data pipeline. You start by suppling an idea for a category of t-shirt. When the pipeline runs, it does a number of things:

• generates p number of patterns using the supplied idea
• generates an image with a transparent background with the generated text
• uploads these images to github
• lists and publishes all the new patterns on printify

From there, Printify takes over and the following things happen

• products are synced with shopify, and a number of other retailers
• orders are automatically sent to the print provider when a sale is made

Running Locally

There are two ways this can be run. The first is as a command line utility, and the second is as a web service. to run it as a command line utility, run the following command:

python3.12 product_pipeline.py -p 10 "Idea for a t-shirt"

In order to run the application as a web service, run the following command:

uvicorn product_pipeline:app --host localhost --port 8080 --reload

Deployment

Google Cloud Functions

Stay tuned for the deployment to GCP!

Reuqired Environment Variables

Run the following command to generate the required .env file. Replace the empty strings with the appropriate values.

cat <<EOF > .env
PRINTIFY_API_KEY=""
OPENAI_API_KEY=""
GH_PAT=""
GH_UPLOAD_REPO=""
GH_CONTENT_PREFIX=""
EOF

Additionally, the shopify utility requires these envronment variables to be set:

echo 'SHOPIFY_API_KEY=""' >> .env
echo 'SHOPIFY_API_ACCESS_TOKEN=""' >> .env
echo 'SHOPIFY_SHOP_NAME=""' >> .env

Running Tests

Formal unit tests are not yet implemented. However, all the functions have a manual test if you run the script directly.

Main Project

clear && flake8 --ignore=E501 --exclude=.venv

FastAPI Component

curl -X POST "http://localhost:8080/process_patterns" \
-H "Content-Type: application/json" \
-d '{"patterns": 1, "idea": "unit testing"}'

Image Utility

python3.12 image_util.py

AI Utility

python3.12 ai_util.py