Turn your python code as a package to be installed everywhere.
Keep track of your different training iterations with MLFlow.
Use the power of the cloud computing to train your model on a very large dataset with Google Colab and AI platform.
- Build your predictive API with FastAPI
- Mount the API in a Docker container
- Deploy the API on Google Cloud Run to make prediction in production
- Build a user interface with Streamlit
- Plug your API to a user interface to build a predictive product
- Deploy a user interface on Heroku for your predictive product