https://github.com/hiborn4/invoiceiq
AI-powered Invoice Data Extraction Tool – Upload any document (PDF/image) and extract key information in structured JSON format using OCR and vision-based LLM techniques, including robust stamp detection.
https://github.com/hiborn4/invoiceiq
Last synced: 12 months ago
JSON representation
AI-powered Invoice Data Extraction Tool – Upload any document (PDF/image) and extract key information in structured JSON format using OCR and vision-based LLM techniques, including robust stamp detection.
- Host: GitHub
- URL: https://github.com/hiborn4/invoiceiq
- Owner: HiBorn4
- Created: 2025-08-06T10:08:43.000Z (12 months ago)
- Default Branch: main
- Last Pushed: 2025-08-06T10:17:29.000Z (12 months ago)
- Last Synced: 2025-08-06T12:10:49.167Z (12 months ago)
- Language: Python
- Size: 7.72 MB
- Stars: 0
- Watchers: 0
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: READme.md
Awesome Lists containing this project
README
# 📄 InvoiceIQ – Intelligent Invoice Data Extraction System
---
## 🧠 Project Overview
**InvoiceIQ** is a powerful document intelligence system that performs **OCR-based data extraction** from uploaded invoices and other structured documents. It reliably parses key information such as:
- Invoice Number
- Invoice Date
- Vendor Code
- PO Number
- Item Details (Name, Quantity, Rate)
- Total & Grand Total
- Stamp Verification (Highly accurate)
The system is designed to **work with diverse layouts** and file formats, extracting data even from **complex or noisy documents**. A major challenge of stamp-based verification has also been successfully tackled using robust **image embedding comparisons**.
Whether the file is a scan, image, or PDF—if a human can read it, **InvoiceIQ** can parse it.
---
## 🚀 Getting Started
### 🖥 Frontend (Streamlit UI)
```bash
streamlit run app.py
````
### 🛠 Backend (FastAPI + Embedding + Utils)
```bash
# Step 1: Install requirements
pip install -r requirements.txt
# Step 2: Start the backend server
cd backend
uvicorn app:app --reload
```
The backend provides:
* OCR processing
* Image preprocessing
* Embedding-based stamp comparison
* JSON output formatting
---
## 🎥 Demo Video
⬇️ *(Demo coming soon here)*
https://github.com/user-attachments/assets/3c9cdbfc-f0f0-422c-83ac-b047b23a81b6
---
## 🔍 Features
✅ Human-readable document parsing
✅ Highly flexible input support (PDF, PNG, JPG, etc.)
✅ Strict positional & format constraints for accuracy
✅ Fast and scalable backend using FastAPI
✅ Accurate stamp detection with `text-embedding-ada-002`
---
## 🛠 Tech Stack
* **Frontend**: Streamlit
* **Backend**: FastAPI
* **OCR Engine**: Tesseract
* **Embedding**: Azure OpenAI `text-embedding-ada-002`
* **Stamp Matching**: Vector similarity + base64 image handling
* **File Types Supported**: PDF, PNG, JPG
---
## 📦 Folder Structure
```
InvoiceIQ/
│
├── app.py # Streamlit Frontend
├── backend/
│ ├── app.py # FastAPI backend
│ ├── utils/ # OCR, Preprocessing, Embedding scripts
│ └── model/ # Embedding logic, Stamp comparison
│
├── public/
│ └── invoice_ui_1.png
│ └── invoice_ui_2.png
│
└── requirements.txt
```
---
## 🌍 Use Cases
* Automated invoice digitization
* Bill processing systems
* Procurement systems
* KYC or stamp-based validation systems
* Document archiving with searchable metadata
---
## 💼 Ideal For
✅ Freelance project delivery on **Upwork** / **Fiverr**
✅ AI / OCR portfolios
✅ Enterprise document processing workflows
---
## 📬 Contact
Feel free to reach out if you'd like a custom solution or integration!
---