Credo - Credit Score API v1.0.0

A FastAPI-based Credit Score Prediction API powered by TensorFlow Machine Learning Models

Features • Quick Start • Usage Examples • API Documentation • Roadmap

---

Overview

Credo is a modern, scalable Credit Score API designed specifically for Fintech applications. It leverages machine learning models built with TensorFlow to predict credit scores based on financial history and application data. The API is built with FastAPI, providing high performance, automatic documentation, and easy integration.

Key Highlights

• High Performance - Built on FastAPI for lightning-fast inference
• ML-Powered - ML models for accurate credit predictions
• Easy Integration - RESTful API with automatic Swagger/OpenAPI documentation
• Production-Ready - Designed with Fintech requirements in mind
• Scalable - Ready for deployment in cloud environments

---

Tech Stack

• Backend Framework: FastAPI
• Python Version: 3.9+
• Database: Sqlite (SQLModel)
• ML Framework: TensorFlow / Keras
• Package Manager: UV

---

Installation

Prerequisites

• Python 3.10 or higher
• UV package manager

Setup Instructions

1. Clone the repository

git clone https://github.com/MarioCanudas/credo.git
cd credo

2. Install dependencies with UV

uv sync

3. Activate virtual environment

source .venv/bin/activate  # On Windows: .venv\Scripts\activate

Start the API

uv run fastapi dev app/main.py

The API will be available at http://127.0.0.1:8000.

---

Quick Start

Once the API is running, you can start making requests. The typical workflow is:

1. Create a User with their financial profile.
2. Submit a Credit Application for that user to get a prediction.

1. Create a User

curl -X POST "http://127.0.0.1:8000/users/" \
     -H "Content-Type: application/json" \
     -d '{
           "gender": "m",
           "car": true,
           "realty": true,
           "cnt_children": 0,
           "income": 200000,
           "income_type": "working",
           "education_type": "higher_education",
           "fam_status": "married",
           "housing_type": "house",
           "mobile_phone": true,
           "work_phone": true,
           "phone": true,
           "email": true,
           "job_title": "managers",
           "cnt_fam_members": 2,
           "age": 35,
           "work_experience": 10,
           "bad_debt": 0,
           "good_debt": 5
         }'

2. Request Credit Prediction

curl -X POST "http://127.0.0.1:8000/applications/" \
     -H "Content-Type: application/json" \
     -d '{"user_id": 1}'

---

API Documentation

Credo provides automatic documentation via Swagger UI. Once the server is running, navigate to:

• Swagger UI: http://127.0.0.1:8000/docs
• ReDoc: http://127.0.0.1:8000/redoc

Core Endpoints

Method	Endpoint	Description
POST	/users/	Create a new user profile
GET	/users/{user_id}	Retrieve user profile details
PUT	/users/{user_id}	Update an existing user profile
DELETE	/users/{user_id}	Delete a user and their data
POST	/applications/	Submit a credit application (Predicts status)
GET	/applications/{user_id}	Retrieve the latest application status

---

Usage Examples

Updating User Information

If a user's income or job status changes, you can update their profile:

curl -X PUT "http://127.0.0.1:8000/users/1" \
     -H "Content-Type: application/json" \
     -d '{
           "gender": "m",
           "car": true,
           "realty": true,
           "cnt_children": 1,
           "income": 250000,
           "income_type": "working",
           "education_type": "higher_education",
           "fam_status": "married",
           "housing_type": "house",
           "mobile_phone": true,
           "work_phone": true,
           "phone": true,
           "email": true,
           "job_title": "managers",
           "cnt_fam_members": 3,
           "age": 36,
           "work_experience": 11,
           "bad_debt": 0,
           "good_debt": 6
         }'

Checking Application Status

curl -X GET "http://127.0.0.1:8000/applications/1"

---

Project Structure

credo/
├── app/                          # FastAPI application
│   ├── main.py                  # Application entry point
│   ├── api/                     # API routes/endpoints
│   ├── core/                    # Core configuration
│   ├── db/                      # Database setup
│   ├── models/                 # Pydantic models
│   ├── services/                 # Business logic and ML integration
│   └── enums/                 # Enums datatypes to be used in the database
├── ml_engine/                    # Machine Learning module
│   ├── artifacts/               # Saved ML models
│   └── training/
│       └── data/               # Training datasets
├── tests/                    # Unit and integration tests
├── data.db                    # Sqlite database file
├── pyproject.toml
└── README.md

---

Roadmap

Feature	Status	Priority	Notes
Initial Project Setup	✅ Completed	🟢 Low Priority	Basic FastAPI structure with UV setup
Design Initial ML Model	✅ Completed	🔴 High Priority	Define features and architecture for the credit score model
Train Initial ML Model	✅ Completed	🔴 High Priority	Develop and train the first TensorFlow model (using Keras)
Determine Basic Schemas	✅ Completed	🟡 Medium Priority	Define basic Pydantic schemas for applications and user info
Implement DataBase	✅ Completed	🔴 High Priority	Implement the database schema and initialization
Implement API Endpoints	✅ Completed	🔴 High Priority	Create the API endpoints to manage petitions and responses
Integrate ML Model with API	✅ Completed	🔴 High Priority	Connect the trained model to the FastAPI endpoints for inference
Testing & Validation	✅ Completed	🟢 Low Priority	Implement unit and integration tests for API and ML model
Documentation & Examples	✅ Completed	🟡 Medium Priority	Create comprehensive documentation and usage examples
Other fancy features	⏳ Planned	🟢 Low Priority	Additional features

Status Legend:

• ✅ Completed
• ⏳ Planned
• 🔄 In Progress
• 🔀 Changed

Priority Legend:

• 🔴 High Priority
• 🟡 Medium Priority
• 🟢 Low Priority