Spending Manager | Extract Payments from Gmail -> Parse -> Add to Sheets

Spending Manager | Extract Payments from Gmail -> Parse -> Add to Sheets Workflow Overview This workflow automates personal expense tracking by processing financial emails from Gmail, extracting transaction details using AI, and logging them into a Google Sheet. It's designed to handle different types of financial emails—such as individual transaction alerts, daily spending summaries, and invoices—by intelligently routing them through different processing paths. The automation eliminates the manual, error-prone task of sifting through emails and copying/pasting spending data into a spreadsheet. By using AI to understand unstructured email content, it provides a sophisticated, hands-off system for maintaining a real-time, categorized log of all expenditures. Technologies Used - Gmail: Acts as the trigger, initiating the workflow whenever a new email with a specific label (e.g., "invoice" or "payment") arrives. - Google Sheets: Serves as the final database, where all extracted and structured transaction data is appended for tracking and analysis. - LangChain: A framework used to build applications with Large Language Models (LLMs). Here, it chains together prompts, AI models, and structured output parsers. - Google Gemini & Groq: The AI models (LLMs) responsible for reading the email content and extracting the relevant financial data. - n8n Native Nodes: - File Extractor: To pull text from PDF attachments like invoices. - HTML Extractor: To parse specific data from the HTML body of emails. - Logic Nodes (Switch, Set, Merge): To route data, standardize formats, and manage different processing flows within the workflow. AI Integration AI is the core intelligence of this workflow, transforming it from a simple data-mover to a sophisticated data-parser. It is integrated via the LangChain nodes to analyze unstructured email content and extract structured financial data. 1. Contextual Analysis: The workflow first categorizes emails based on the sender. This context (e.g., "this is a daily summary from Cathay Bank" or "this is a single crypto.com transaction") is passed to the AI. 2.