Automated Financial Document Processing with Google Gemini OCR — n8n 워크플로

높음 복잡도 트리거35개의 노드🏷️ Invoice Processing작성자: Didac Fernandez

개요

AI-Powered Financial Document Processing with Google Gemini

This comprehensive workflow automates the complete financial document processing pipeline using AI. Upload invoices via chat, drop expense receipts into a folder, or add bank statements - the system automatically extracts, categorizes, and organizes all your financial data into structured Google Sheets.

What this workflow does

Processes three types of financial documents automatically:

Invoice Processing**: Upload PDF invoices throu

사용된 노드

Google SheetsHTTP RequestGoogle DriveCodeAI AgentBasic LLM ChainStructured Output ParserOpenRouter Chat Model

워크플로 미리보기

Mark - Accountant
!My Image
Invoice Processing
Expenses Processing
Donna - Accountant
!My Image
Expenses Processing
Victor - Controller
!My Image
Andrew - CFO
!My Image
Finance
AutoSolutions.ai - AI Consulting Services
Didac Fernandez Girona
!My Image
🤖 AI-Powered Financial Document Processing W
📋 Overview
This comprehensive workflow automates financial documen
🛠️ Setup Requirements
🌐 Google Services
1. 🚗 Google Drive API credentials configured
2. 📊 Google Sheets API credentials configured
3. 🧠 Google Gemini API credentials configured
4.
🎯 How to Use
📄 Processing Invoices
1. 💬 Access the chat interface
2. 📤 Upload PDF invoice files
3. ⏳ Wait for AI processing and confirmation
4. ✅ Check "Test Invoice Records" Google Sheet for e
📄 INVOICE PROCESSING SECTION
🎯 Purpose
This section handles invoice processing through a chat
🔄
Mark - Accountant 👨‍💼
Role: Invoice Processing Specialist
Responsibilities:
- ✅ Extract vendor information
- ✅ Capture payment terms
- ✅ Process line item details
📤 STEP 1: File Upload & Storage
Purpose: Save uploaded invoice to Google Drive
Configuration:
- 📁 Target Folder: "Invoices"
- 📄 Input Field: data0 (from chat upload)
- 📝 Filename: Uses ori
🤖 STEP 2: AI Upload
Purpose: Upload PDF to Google Gemini for processing
Configuration:
- 🔗 Endpoint: Gemini File Upload API
- 📄 Content-Type: application/pdf
- 🔐 Auth: Google PaLM API cred
🧠 STEP 3: Data Extraction
Purpose: Use Gemini AI to extract structured invoice da
AI Prompt: Extracts:
- 🏢 Vendor Name
- 🔢 Invoice Number
- 📅 Invoice Date
📋 STEP 4: JSON Parsing
Purpose: Clean and parse AI response into structured da
Process:
- 🧹 Remove markdown code blocks
- 📊 Convert text to JSON object
- ⚠️ Handle parsing errors gracefu
📑 STEP 5: Line Item Processing
Purpose: Split line items for individual spreadsheet ro
Process:
- 📋 Extract Line Items array from parsed data
- 🔄 Create separate execution for each line
📊 STEP 6: Save to Spreadsheet
Purpose: Store extracted invoice data in Google Sheets
Mapping:
- 🏢 Vendor Name → Header data
- 🔢 Invoice Number → Header data
- 📅 Dates → Header data
🔗 STEP 7: Data Merge
Purpose: Combine file metadata with processed data
Inputs:
- 📁 File information from Google Drive
- 📊 Processed invoice data from sheets
Output: Combined data for
📝 STEP 8: File Renaming
Purpose: Rename file with vendor and invoice number
Format: "{Vendor Name} - {Invoice Number}"
Example: "Acme Corp - INV-2024-001"
Error Handling: Continu
✅ STEP 9: Confirmation
Purpose: Send success message to user
Message: Confirms invoice processing completion
Includes: Reference to check Google Sheet
User Experience: Clear feed
🧾 EXPENSE PROCESSING SECTION
🎯 Purpose
This section monitors a Google Drive folder for new exp
📁 EXPENSE TRIGGER
Purpose: Monitor Google Drive folder for new expense re
Configuration:
- 📂 Folder: "Expense Receipts"
- ⏰ Frequency: Every minute
- 🚨 Event: File created
🏷️ AI CATEGORIZATION
Purpose: Automatically categorize expenses using AI
Process:
- 🤖 Uses OpenRouter LLM
- 📋 17 predefined categories
- 🎯 Structured JSON output
🏦 BANK STATEMENT PROCESSING SECTION
🎯 Purpose
This section monitors a Google Drive folder for new ban
Victor - Controller 👨‍💼
Role: Bank Statement Processing & Transaction Reconcili
Responsibilities:
- ✅ Extract transaction details
- ✅ Process multiple transactions per statement
- ✅ M
🔄 CUSTOM TRANSACTION PARSER
Purpose: Parse multiple bank transactions from AI respo
Features:
- 🧹 Cleans JSON formatting
- 📊 Handles multiple response formats
- 🔄 Converts to n8n array
📈 FINANCIAL ANALYSIS AGENT SECTION
🎯 Purpose
Andrew, the AI CFO, provides intelligent financial anal
Andrew - CFO 👨‍💼
Role: Financial Analysis & Strategic Reporting
Responsibilities:
- ✅ Generate financial insights
- ✅ Answer complex queries
- ✅ Create executive reports
📊 ANALYSIS TRIGGER
Purpose: Manual trigger for financial analysis queries
Usage:
- 🔘 Enable this trigger when analysis is needed
- 💬 Send financial queries to the AI agent
- 📈 Get insights
🗃️ DATA ACCESS TOOLS
Purpose: Connect AI agent to financial spreadsheets
Tools Available:
- 📄 Invoices Tool: Invoice records and billing data
- 🧾 Expenses Tool: Categorized expense
Donna - Accountant 👩‍💼
Role: Expense Processing & Categorization Specialist
Responsibilities:
- ✅ Extract merchant & transaction details
- ✅ Categorize expenses automatically
- ✅ Process r
!My Image
!My Image
Virtual Accounting Department
tooltooltoolmodelmodelparser
W
When chat message received
S
Split Out
C
Confirmation
P
Parsed Data
Upload PDF to Google Gem…
Download Data from Googl…
Table ERP
W
When Executed by Another…
M
Merge
Update File Name
Save Invoice
G
Google Drive Trigger
S
Split Out1
P
Parsed Data1
Upload PDF to Google Gem…
Download Data from Googl…
Google Sheets
Google Drive
G
Google Drive Trigger1
Upload PDF to Google Gem…
Download Data from Googl…
Google Sheets1
Google Drive1
E
Edit Fields
E
Edit Fields1
Code
Basic LLM Chain
OpenRouter Chat Model
Structured Output Parser
AI Agent
OpenRouter Chat Model1
I
Invoices
E
Expenses
T
Transactions
E
Edit Fields2
35 nodes35 edges

작동 원리

  1. 1

    트리거

    워크플로는 트리거 트리거로 시작합니다.

  2. 2

    처리

    데이터가 35개의 노드를 통해 흐릅니다, connecting agent, chainllm, chattrigger.

  3. 3

    출력

    워크플로가 자동화를 완료하고 구성된 대상에 결과를 전달합니다.

노드 세부 정보 (35)

GO

Google Sheets

googleSheets

#1
HT

HTTP Request

httpRequest

#2
GO

Google Drive

googleDrive

#3
CO

Code

code

#4
AI

AI Agent

n8n-nodes-langchain.agent

#5
BA

Basic LLM Chain

n8n-nodes-langchain.chainLlm

#6
ST

Structured Output Parser

n8n-nodes-langchain.outputParserStructured

#7
OP

OpenRouter Chat Model

n8n-nodes-langchain.lmChatOpenRouter

#8

이 워크플로 가져오는 방법

  1. 1오른쪽의 JSON 다운로드 버튼을 클릭하여 워크플로 파일을 저장합니다.
  2. 2n8n 인스턴스를 열고 워크플로 → 새로 만들기 → 파일에서 가져오기로 이동합니다.
  3. 3다운로드된 automated-financial-document-processing-with-google-gemini-ocr 파일을 선택하고 가져오기를 클릭합니다.
  4. 4각 서비스 노드에 대한 자격 증명(API 키, OAuth 등)을 설정합니다.
  5. 5워크플로 테스트를 클릭하여 모든 것이 작동하는지 확인한 후 활성화합니다.

또는 n8n → JSON에서 가져오기에 직접 붙여넣기:

{ "name": "Automated Financial Document Processing with Google Gemini OCR", "nodes": [...], ...}

통합

agentchainllmchattriggercodeexecuteworkflowtriggergoogledrivegoogledrivetriggergooglesheetsgooglesheetstoolhttprequestlmchatopenroutermergeoutputparserstructuredsetsplitout

이 워크플로 가져오기

한 번의 클릭으로 다운로드 및 가져오기

JSON 다운로드n8n.io에서 보기
노드35
복잡도high
트리거trigger
카테고리Invoice Processing

제작자

Didac Fernandez

Didac Fernandez

@didac-fg

태그

agentchainllmchattriggercodeexecuteworkflowtriggergoogledrivegoogledrivetriggergooglesheetsgooglesheetstoolhttprequest

n8n을 처음 사용하시나요?

n8n은 무료 오픈소스 워크플로 자동화 도구입니다. 자체 호스팅하거나 클라우드 버전을 사용하세요.

n8n 무료로 시작하기 →

Related Invoice Processing Workflows

COCOEMEX+5
medium

Automate Custom QuickBooks Invoice PDFs & Email with n8n

Standard accounting templates often fail to reflect a premium brand identity. This sophisticated n8n workflow bridges the gap between financial record-keeping and professional client presentation. By moving beyond the native limitations of QuickBooks Online, this automation enables businesses to generate high-end, multi-page PDF invoices that align perfectly with their corporate styling. The process begins the moment a new invoice is generated in QuickBooks, triggering a webhook that captures real-time billing data. The workflow then utilizes advanced HTML-to-File conversion and custom Code nodes to structure data into a polished, branded layout. It handles complex logic such as line-item merging and multi-page formatting automatically. Once the document is rendered, the system bypasses generic 'no-reply' senders by routing the finalized PDF through your preferred email provider. This ensures a seamless, white-labeled experience for your clients while eliminating the manual overhead of exporting, styling, and attaching files. Ideal for agencies and service providers, this flow guarantees that your most frequent touchpoint—the bill—is as professional as your work. **Common Use Cases:** - High-end creative agencies requiring bespoke, white-labeled billing documents for premium clients. - Automated recurring subscription billing where custom tax disclosures or localized branding are required. - Service-based businesses needing to attach dynamic project reports or terms of service directly to QuickBooks invoices.

🔗 Webhook·12 nodes