azf99/bank-statement-analysis

Extract useful insights from PDF Bank Statements(Indian Banks) using python automation

Hi 👋, I'm Azfar Lari

A passionate ML Developer, previously active on Upwork(Top Rated)

• 🌱 I’m currently learning Django, Google Cloud, NLP

• 👨‍💻 My Portfolio on Upwork here

• 💬 Ask me about TensorRT, MLOps, Python, OpenCV

• 📫 How to reach me [email protected]

• 📄 Know about my experiences here.

Connect with me:

Languages and Tools:

 

bank-statement-analysis's Issues

Add interactive dashboard

For better displaying results of the analyses

Increased Bank Support

I was working on a project for which I found your repository very useful, I wanted to increase the number of banks that you recognised, It'd be very useful if you gave me a few insights of your code in order for me to modify it for other banks as I am facing a few setbacks. A readme file/Email reply would be much appreciated which will help me to modify your code,Once I do so I will share it with you. It'd be much appreciated.

Improve transaction classification(Maybe NLP?)

Better classification of bank transactions

Help international use

Could you please put all specific data in a single place json file?
e.g.:

banks = {"YES": "YES BANK", "ALLA": "ALLAHABAD BANK"}
title = ["mr.", "shri", "ms.", "mrs."]
df = df[["Transaction Date", "Value Date", "Description", "Debit", "Credit", "Balance"]]
"%d/%m/%Y
test = pd.DataFrame(columns=["Transaction Date", "day", "month", "year", "Balance"], index=df.index)
"Average monthly debit = Rs."
labels = {"imps": "imps", "rrn": "imps", "loan": "loan", "emi": "emi", "amazon": "shopping", "flipkart": "shopping",
              "mutualfund": "invest", "txn paytm": "trf", "restaurant": "food", "paytm": "trf",
              "atd": "atm", "atm": "atm", "net txn": "nettxn", "cash": "cash", "funds trf": "trf", "neft": "neft",
              "interest": "interest",
              "metro": "travel", "swiggy": "food", "faasos": "food", "zomato": "food", "upi": "trf", "ola": "travel",
              "refund": "refund",
              "charge": "bank_charges", "pca": "trf"}

Projects

Add parellization to data extraction from pages

Convert tabula and camelot to PyPDF2 or pdfminer.six

Currently, using wrapper classes is very slow for data extraction