Skip to content

Latest commit

 

History

History

README.md

🏪 Credit Risk Model with Feature Store

This project demonstrates a Credit Risk Assessment Model using JFrogML's Feature Store - from feature engineering to model deployment with real-time inference.

📋 Prerequisites


🚀 JFrogML Feature Store Workflow

Sequential workflow - Phase 2 depends on Phase 1 completion:

🏪 Phase 1: Feature Store Setup (Required First)

┌─────────────────┐    ┌─────────────────┐    
│   📊 Data       │ -> │   🔧 Feature    │ 
│   Source        │    │   Set           │
│   Registration  │    │   Registration  │
└─────────────────┘    └─────────────────┘

Complete workflow: 🏪 Feature Store Setup & Testing Guide

Purpose: Set up data connectors, feature transformations, and validate Feature Store components



🚀 Phase 2: Model Training & Deployment (Depends on Phase 1)

┌─────────────────┐    ┌─────────────────┐    ┌─────────────────┐
│   💻 ML App     │ -> │   🏗️ Build      │ -> │   🚀 Deploy     │
│   Code          │    │   (Training)    │    │   ML Serving    │
│   + Features    │    │   Offline Store │    │   Online Store  │
└─────────────────┘    └─────────────────┘    └─────────────────┘

Complete workflow: 🚀 Model Training & Deployment Guide

Purpose: Build and deploy ML models using the registered Feature Store components



📁 Project Structure

feature_set_quickstart_guide/
|
├── README.md                    # This overview guide
├── feature_store/               # Feature Store components
│   ├── data_source.py           # Data connector (S3 CSV)
│   └── feature_set.py           # Feature transformations
|
├── main/                        # ML model code
│   ├── __init__.py              # Python package initialization
│   ├── model.py                 # CatBoost credit risk model
│   ├── utils.py                 # Data utilities
│   └── conda.yaml               # Environment dependencies
|
├── feature-store-setup.md       # 🏪 Phase 1 guide
└── model-training-deployment.md # 🚀 Phase 2 guide