nvbinh15 / credit-assessment-microservice

credit-assessment-microservice

Decription

This repo contains source code for a loan management and credit assessment microservice powered by machine learning models, which was submitted to NUS Fintech Month 2023 Hackathon.

Pitch Deck: Canva

Demo: Youtube

Deployment

The microservice is deployed on an AWS EC2 instance with an Nginx reverse proxy.

Deployment: http://18.142.225.55:8000/docs

Features

Public-available API endpoints:

• Credit assessment: classify user’s credit score into 3 different groups: Good, Standard, and Poor based on customers’ information
• Loan prediction: predict whether a personal loan will be accepted or not
• SME loan estimation: calculate the amount of expected loan that SMEs will receive based on market-standard collateral analysis

Loan management API: serves as the backend for a loan management mobile application.

Getting started

• Create and activate a virtual environment
• Install dependencies: $ pip install -r requirements.txt
• Create an environment file app/.env with a SECRET_KEY and activate the file by $ source app/.env
• Start the microservice: uvicorn app.main:app --reload