Building a Local Receipt Extractor with Open-Source AI (Side Hustle)

This is a submission for the Hacktoberfest Open-Source AI Challenge Week 2: Side Hustle.

What I Built

I built a Local Receipt & Invoice Extractor. It is a privacy-first tool for freelancers and small business owners. You upload a receipt image, and it extracts the Vendor Name, Date, Total Amount, and Tax Amount. The biggest problem it solves is data privacy—freelancers don't want to upload sensitive client bills to cloud servers like Google or OpenAI.

Demo

·

How I Built It

· UI: Streamlit (Python)
· OCR (Image to Text): Tesseract OCR
· AI Parsing: Ollama running the Llama 3 model locally.

The app reads the image using Tesseract, passes the raw text to the local Llama 3 model, and extracts structured data.

Why Does Open Innovation Matter?

This is where open-source shines:

1. 100% Offline & Private: Because I used Ollama to run Llama 3 locally, the user's financial data never leaves their laptop. No API calls, no cloud.
2. Zero API Costs: Cloud AI APIs charge per token. By running the model locally, the app costs $0 to run. This makes it a viable side hustle without monthly bills.
3. No Lock-in: I can swap models or fine-tune prompts for Indian-style GST invoices without being locked into a provider's pricing.

My First User

I tested this with a friend who runs a small boutique. They currently keep their bills in a shoebox. They said they would absolutely pay ₹200/month for a simple tool that sorts their bills securely without uploading them to the cloud.