FedEx Flex-Pay AI 🚚💸

An Intelligent Billing Agent designed to optimize the "Cost to Collect" while increasing "Customer Lifetime Value".

🚀 Overview

FedEx Flex-Pay AI transforms the traditional billing process from a transactional demand into a relationship-building opportunity. By leveraging a Risk Scoring Model and contextual signals (like anniversaries or payment history), specifically targeted at reducing churn and improving collection rates without human intervention.

Key Value Proposition

• 📉 Reduce Cost to Collect: Automated, empathetic AI agents handle low-to-mid risk accounts, saving agent time for critical cases.
• 📈 Increase CLV: "Anniversary Forgiveness" and "Good Faith Discounts" turn friction points into loyalty moments.

---

🧠 The AI Model & Pipeline

The system operates on a 4-stage pipeline defined in ARCHITECTURE.md.

1. Ingestion: Parses input_data.json for account lifecycle data.
2. Risk Scoring (0-100):
   • Low Risk (<30): Eligible for "Early Bird" discounts.
   • High Risk (>70): Flagged for empathetic but firm reminders.
   • Formula: Base(50) + HistoryFactor + BalanceFactor
3. Decision Engine: Selects the optimal retention strategy (Waiver vs. Discount vs. Reminder).
4. Presentation: Delivers the tailored message via the Admin Dashboard or Member Portal.

---

🛠️ Tech Stack

• Backend: Python 3.9+ (Logic, Risk Model, Text Generation)
• Frontend: HTML5, CSS3 (FedEx Brand Identity), Vanilla JS
• Data: JSON (No SQL DB required for this prototype)
• Security: XSS-Hardened Frontend, Input Validation

---

📂 Project Structure

├── assets/
│   ├── css/style.css       # FedEx Brand Styling
│   ├── js/script.js        # Admin Dashboard Logic (Risk Viz)
│   ├── js/member.js        # Member Portal Logic
│   ├── js/auth.js          # Mock Authentication
│   └── js/data.js          # Shared Account Data
├── backend/
│   ├── flex_pay.py         # Core Python AI Logic
│   ├── tests.py            # Unit Tests for Risk/Discounts
│   └── input_data.json     # Seed Data
├── index.html              # Admin Dashboard (Main View)
├── login.html              # Login Screen (Entry Point)
├── member.html             # Member Portal (User View)
└── ARCHITECTURE.md         # Detailed System Design

---

📥 Installation & Setup

To run this project locally, follow these simple steps:

1. Clone the Repository

   git clone https://github.com/GVN2307/Fedx-Flex-pay-AI.git
   cd Fedx-Flex-pay-AI

2. No Dependencies Required

   • The project uses standard Python (3.x) and vanilla HTML/JS. No pip install or npm install needed!

---

🚦 How to Run

1. Web Interface (Recommended)

Open login.html in any modern web browser.

Logins:

• Manager View:
  • Username: admin
  • Password: admin
  • View: Overall Business Impact, Risk Analysis of all accounts.
• Member View:
  • Username: A001 (or A002, A003...)
  • Password: password
  • View: Personalized offer specific to that customer.

2. Backend Logic (CLI)

To test the Python message generation engine directly:

cd backend
python flex_pay.py

3. Run Tests

Verify the Risk Model and Discount Logic:

cd backend
python tests.py

---

🔒 Security

This application has been audited for:

• XSS (Cross-Site Scripting): All user inputs and data rendering are sanitized using escapeHtml().
• Injection Attacks: Data is handled via structured JSON parsing, mitigating SQLi risks.

---

📜 License

This project is a prototype developed for the FedEx Flex-Pay AI initiative.