ArticWasHere/data-entry-warehouse

DataEntry AI — Bank & Warehouse Document Extraction

An AI model fine-tuned on DistilBERT to automatically extract key fields from bank and warehouse documents. Built for B2B licensing to banks and warehousing companies.

What it does

Paste any bank or warehouse document and it instantly extracts structured fields — no manual data entry needed.

Bank Documents

Extracts: ACCOUNT_NUMBER · ACCOUNT_HOLDER · DATE · TRANSACTION_TYPE · AMOUNT

Warehouse / Invoice Documents

Extracts: INVOICE_NUMBER · VENDOR · ITEM_NAME · QUANTITY · UNIT_PRICE · TOTAL_AMOUNT

---

Quick Start

from transformers import pipeline

# Bank documents
bank_model = pipeline("token-classification", model="ArticWasHere/data-entry-bank")
result = bank_model("Account Number: 4821039476 Amount: $4,250.00 Date: 03/15/2024")
print(result)

# Warehouse documents
warehouse_model = pipeline("token-classification", model="ArticWasHere/data-entry-warehouse")
result = warehouse_model("Invoice INV-20481 Vendor: Apex Supply Co Quantity: 12 units")
print(result)

---

Example Input & Output

Bank Document Input

Account Number: 4821039476
Account Holder: Sarah Johnson
Date: 03/15/2024
Transaction Type: deposit
Amount: $4,250.00

Extracted Output

{
  "ACCOUNT_NUMBER": "4821039476",
  "ACCOUNT_HOLDER": "Sarah Johnson",
  "DATE": "03/15/2024",
  "TRANSACTION_TYPE": "deposit",
  "AMOUNT": "$4,250.00"
}

---

Model Details

Property	Value
Base Model	distilbert-base-uncased
Task	Token Classification (NER)
Training Samples	10 labeled documents
Epochs	5
Languages	English

---

Intended Use

• Automating data entry for banks and financial institutions
• Processing warehouse invoices and purchase orders
• Reducing manual data entry errors in enterprise workflows

---

Roadmap

• Add more training data (100+ samples)
• Support for more document types (KYC, loan applications)
• Multi-language support
• REST API endpoint

---

Built by ArticWasHere · GitHub · Contact for licensing inquiries