Project

Zedl

Visit the site ↗
FastAPI SQLAlchemy Google Document AI Gemini DATEV/BMD PWA

Problem

Dental practices, freelancers, and sole proprietors still enter receipts by hand for the most part. That costs time, invites errors, and handing the data over to a tax advisor or accounting software stays cumbersome. Manually entered invoice data leads to mistakes, and without a consistent structure, archiving quickly becomes hard to manage.

Solution

  • → AI-based text recognition using Google Document AI and Gemini
  • → Automatic extraction of date, amount, tax rates, supplier, and line items
  • → Export to DATEV, BMD, CSV, and Excel for direct use in accounting software
  • → Progressive Web App for scanning and capture straight from a smartphone
  • → Intelligent duplicate detection prevents double entries
  • → Automatic currency conversion at ECB rates for international receipts
  • → GoBD- and BAO-compliant encrypted archiving for 10 years

Result

Zedl takes over the entire manual receipt-entry process. Receipts are scanned, categorized automatically, and delivered in a format that imports directly into DATEV, BMD, or other accounting software.

The AI reliably identifies every relevant data point on receipts and invoices: line items are extracted automatically, tax rates recognized, receipts categorized intelligently. The mobile PWA supports scanning on site, while the web app provides full management and export capabilities.

Key features

AI text recognition

Google Document AI and Gemini combined for precise recognition

Line-item extraction

Automatic recognition of individual line items with quantities, prices, and tax rates

Multi-format export

DATEV, BMD, CSV, and Excel for every common accounting system

Mobile PWA

Installable app for iOS and Android, usable offline

Duplicate detection

Hash-based detection with heuristic checks prevents receipts from being captured twice

Currency conversion

Automatic conversion to euros based on ECB exchange rates

Zedl – screenshot of the receipt scanner

What the customer said

Thank goodness there are sharp minds like yours who work out the perfect solution for problems like this!

Lawrence

Zedl customer

← Back to projects

Facing something similar?

Let's talk