Rene Fichtmueller 2052d87ba1 feat: initial release — AI document intelligence for Paperless-ngx

PaperCortex adds semantic search, auto-classification, receipt extraction,
bank statement matching, and DATEV export to Paperless-ngx — powered
entirely by local AI through Ollama. Exposes everything as an MCP Server
for Claude Code and AI agent integration.

- MCP Server with 5 tools (search, classify, receipt, query, export)
- Local Ollama embeddings for semantic document search
- Receipt data extraction (vendor, amount, date, tax, line items)
- DATEV Buchungsstapel CSV export for German accounting
- Bank CSV transaction matching
- Paperless-ngx REST API client
- Docker deployment
- Zero cloud dependencies — 100% self-hosted

2026-03-26 06:28:48 +13:00

2.6 KiB

Raw Blame History

Receipt Workflow

Overview

PaperCortex provides a complete receipt-to-accounting pipeline:

Scan -- Upload receipts to Paperless-ngx (scan, email, photo)
Extract -- AI extracts structured data (vendor, date, amounts, line items)
Match -- Reconcile against bank CSV exports
Export -- Generate DATEV-compatible CSV for accounting software

Receipt Extraction

Via MCP Server (Claude Code)

Extract receipt data from document #1234

Via CLI

npm run receipt:extract -- --document-id 1234

Extracted Fields

Field	Description	Example
vendor	Company name	"IKEA Deutschland GmbH"
vendorAddress	Full address	"Am Wanderweg 1, 65719 Hofheim"
vendorTaxId	Tax ID / VAT number	"DE 129 341 800"
date	Receipt date	"2024-03-15"
currency	ISO 4217 code	"EUR"
subtotal	Before tax	84.03
taxRate	Tax percentage	19
taxAmount	Tax amount	15.97
totalAmount	Total with tax	100.00
paymentMethod	How it was paid	"card"
lineItems	Individual items	Array of items
category	Expense category	"office_supplies"

Bank Statement Matching

Match receipts against bank CSV exports to verify which receipts correspond to which bank transactions.

Supported Bank Formats

Sparkasse (semicolon-separated, German format)
ING (semicolon-separated)
DKB (semicolon-separated)
Volksbank (semicolon-separated)
Generic CSV

Matching Algorithm

Amount match -- Exact or close amount (within 1.00 tolerance)
Date proximity -- Same day, within 3 days, or within 7 days
Vendor name -- Partial match in transaction description

Results include a confidence score (0.0 - 1.0) and match reasons.

DATEV Export

Format

PaperCortex generates DATEV Buchungsstapel (posting batch) format CSV, compatible with:

DATEV Unternehmen Online
lexoffice
sevDesk
FastBill
Any DATEV-import-capable software

Account Mapping (SKR03)

Category	Account	Description
office_supplies	4930	Buerokosten
travel	4660	Reisekosten
food	4650	Bewirtungskosten
telephone	4920	Telefon
postage	4910	Porto
rent	4210	Miete
advertising	4600	Werbekosten
software	4964	Software
consulting	4950	Rechts- und Beratungskosten
default	4900	Sonstige Aufwendungen

Export via CLI

# Export all receipts from March 2024 as DATEV CSV
npm run receipt:export -- --format datev --year 2024 --month 03

Export via MCP Server

Export documents #100, #101, #102 as DATEV CSV

2.6 KiB Raw Blame History