💸 Invoice Expense Classifier AI

AI-Powered Invoice & Expense Classification System

Automatically classify invoice descriptions into business expense categories using Machine Learning and NLP.

---

📌 Overview

Invoice Expense Classifier AI is a production-ready Machine Learning API that automatically predicts expense categories from invoice text descriptions.

The system uses:

• Natural Language Processing (NLP)
• TF-IDF Vectorization
• Logistic Regression
• FastAPI REST API

This project helps businesses automate financial workflows and reduce manual expense categorization efforts.

---

🚀 Features

✅ Core Features

• AI-powered invoice classification
• FastAPI REST API
• Confidence score prediction
• TF-IDF text vectorization
• Logistic Regression classifier
• Real-time predictions
• JSON API responses

---

⚡ Production Features

• Docker support
• Clean architecture
• Modular codebase
• Swagger API documentation
• Training pipeline included
• Unit testing support
• Lightweight deployment
• Scalable backend design

---

🧠 Supported Categories

The model predicts the following expense categories:

• Logistics
• Office Supplies
• Cloud/Software
• Utilities
• Travel
• Inventory

---

Live Demo link: https://invoice-expense-classifier.onrender.com/docs

---

output screens

Input

output

🧠 Machine Learning Workflow

flowchart TD

    A[Invoice Text]
    --> B[Text Preprocessing]

    B --> C[TF-IDF Vectorization]

    C --> D[Logistic Regression Model]

    D --> E[Expense Category Prediction]

    E --> F[Confidence Score]

Loading

---

📂 Project Structure

Invoice-Expense-Classifier/
│
├── app/
│   ├── main.py
│   ├── model.py
│   ├── preprocessing.py
│   ├── schemas.py
│   └── utils.py
│
├── data/
│   └── train.csv
│
├── models/
│   ├── model.pkl
│   └── vectorizer.pkl
│
├── tests/
│   └── test_api.py
│
├── train.py
├── requirements.txt
├── Dockerfile
├── README.md
└── .gitignore

---

⚙️ Installation & Setup

1️⃣ Clone Repository

git clone https://github.com/Bhavana998/Invoice-Expense-Classifier.git

cd Invoice-Expense-Classifier

---

2️⃣ Create Virtual Environment

Windows

python -m venv venv

venv\Scripts\activate

Linux / Mac

python3 -m venv venv

source venv/bin/activate

---

3️⃣ Install Dependencies

pip install -r requirements.txt

---

🏋️ Train Machine Learning Model

Run the training pipeline:

python train.py

Generated model files:

models/model.pkl
models/vectorizer.pkl

---

🚀 Run FastAPI Server

uvicorn app.main:app --reload

Server URL:

http://127.0.0.1:8000

Swagger API Documentation:

http://127.0.0.1:8000/docs

---

📊 Sample Training Dataset

data/train.csv

text,category
AWS monthly hosting charges,Cloud/Software
Blue Dart courier delivery,Logistics
Printer paper purchase,Office Supplies
Electricity bill payment,Utilities
Flight booking for meeting,Travel
Warehouse inventory materials,Inventory

---

🔌 API Usage

POST /predict

Predict expense category from invoice text.

---

📥 Request

📌 Sample API Requests

Example 1 — Cloud/Software

Request

{
  "text": "Delta flight ticket JFK to LAX for business conference"
}

Response

{
  "category": "Travel",
  "confidence": 0.77
}

---

Example 2 — Logistics

Request

{
  "text": "DHL express courier for warehouse delivery"
}

Response

{
  "category": "Logistics",
  "confidence": 0.95
}

---

Example 3 — Office Supplies

Request

{
  "text": "Staples printer paper ream 500 sheets"
}

Response

{
  "category": "Office Supplies",
  "confidence": 0.94
}

---

Example 4 — Utilities

Request

{
  "text": "Electricity bill for March with demand charges"
}

Response

{
  "category": "Utilities",
  "confidence": 0.96
}

---

Example 5 — Travel

Request

{
  "text": "Delta flight ticket JFK to LAX for business conference"
}

Response

{
  "category": "Travel",
  "confidence": 0.93
}

---

🧪 API Testing Examples

cURL Example

curl -X POST "http://127.0.0.1:8000/predict" \
-H "Content-Type: application/json" \
-d "{\"text\":\"Blue Dart courier charges\"}"

---

Python Request Example

import requests

url = "http://127.0.0.1:8000/predict"

payload = {
    "text": "Electricity bill for warehouse"
}

response = requests.post(url, json=payload)

print(response.json())

---

🐳 Docker Deployment

Build Docker Image

docker build -t invoice-classifier .

---

Run Docker Container

docker run -p 8000:8000 invoice-classifier

---

🧪 Running Tests

pytest

---

📈 Model Details

Component	Technology
NLP Technique	TF-IDF
ML Algorithm	Logistic Regression
Backend Framework	FastAPI
Serialization	Pickle
Testing Framework	Pytest

---

🌍 Deployment Options

The project can be deployed on:

• Render
• Railway
• Docker
• AWS EC2
• Azure
• Google Cloud Platform
• Kubernetes

---

👩‍💻 Author

setty Bhavana

GitHub Profile

https://github.com/Bhavana998

---

⭐ Support

If you found this project useful:

• ⭐ Star the repository
• 🍴 Fork the project
• 🚀 Contribute improvements

---

📜 License

This project is licensed under the MIT License.

---

🚀 Built with FastAPI + Machine Learning + NLP