Firebase + Cloudrun + Machine learning

A simple loan eligibility decision system developed with fastapi and scikit learn for the server side and HTML, CSS and Javascript for the client side

The Backend is deployed on google cloud run while the static frontend is served with firebase hosting.

How to serve the frontend locally

• install firebase tools

npm install -g firebase-tools

• serve on localhost

firebase serve

How to serve the backend locally [Docker]

• navigate to the server folder and create a .env file and enter the following values

IS_DEBUG = False
API_KEY = 
   
    
DEFAULT_MODEL_PATH=./assets/finalized_model.sav

   

you can simply generate an api key using the python repl as follows:

import uuid
print(str(uuid.uuid4()))

• run the shell scripts as follows:

chmod +x run.sh
chmod +x build.sh

• build the image

./build.sh

• serve the api

./run.sh

• navigate to the api documentation on localhost:8080/docs