yeukai08/ibscss

Integrated Behavioural Credit Scoring System (IBCSS)

An AI-driven centralized financial intelligence platform designed to support loan risk assessment by analyzing cross-sector financial transaction behaviour.

Core Features

• —AI Behavioural Scoring: Scikit-learn RandomForest model predicting credit risk.
• —Customer Segmentation: KMeans clustering for borrower categorization.
• —Fraud Detection: IsolationForest + Rule-based anomaly detection.
• —Role-Based Dashboards: Separate interfaces for Admins, Institutions, and Analysts.
• —REST API: Django REST Framework for transaction ingestion.
• —Data Visualization: Rich Chart.js integrations with Tailwind CSS.

Quick Start

1. 1.Install dependencies: pip install -r requirements.txt
2. 2.Start server: python manage.py runserver
3. 3.Visit: http://127.0.0.1:8000

Demo Credentials

• —Admin: admin / admin123
• —Bank: bank1 / bank123
• —Analyst: analyst1 / analyst123

Deployment Notes

• —By default, the app uses SQLITE_PATH (set to /data/db.sqlite3 in Docker Space).
• —If DATABASE_URL is set, it will use that database instead (e.g., PostgreSQL).
• —Startup runs migrations plus seed_demo to ensure demo users and sample records exist.

Implementation Details

• —Backend: Python Django 5.2.x
• —Frontend: Django Templates + Tailwind CSS (CDN)
• —Database: SQLite3
• —AI/ML: Scikit-Learn (RandomForest, KMeans, IsolationForest)