Credit Worthiness Checker

This repository holds programs files required to train and deploy a credit worthiness cheker as an ml service endpoint. The final frontend that uses the backend service can be accessed here credit worthiness checker

This project has been streamlined into 2 main parts:

1. The data science lifecycle - ending with an exported model

2. The deployment - further divided into 2 parts:

   • The backend service
   • The frontend

Requirements

Python 3.8+

Replicating the project

1. Clone this repository to your local machine using the command below:

git clone https://github.com/olumideodetunde/credit_worthiness_checker.git

2. Download the dataset from kaggle and place the data appropriately

   • Click here to download the data & place it in the artifacts/data/raw directory.

3. Run the data science pipeline

   • Run the command leveraging makefile: this would prepare the data, generate defined visualisations, engineer features, train the model and save the model to the deploy directory.

     make machinelearningmodel

4. Deploy the backend service

   • This project used FastAPI to deploy the backend service, To host the backend service, this project leveraged the the dockerfile in the backend directory and follow this link.

5. Build the frontend

   • Leveraging the streamlit script in the frontend directory, you can build the frontend by following the steps found in this link streamlit