Automate LINE Chat with Google Calendar & Gmail Using n8n

Opening Problem Statement

Meet Yuki, a busy event coordinator based in Tokyo. Her daily routine involves juggling back-to-back meetings, emails, and managing her calendar on the go. Yuki often uses LINE for quick communication with clients and teammates, but handling scheduling requests through chat messages can be overwhelming. She spends an average of 3-4 hours per week toggling between her LINE chats, Google Calendar, and Gmail inbox, manually transferring information, which sometimes leads to missed or double-booked appointments.

This manual shuffle causes Yuki stress, wastes valuable time, and occasionally leads to customer dissatisfaction when schedule details get lost or delayed. What if Yuki could simply type her scheduling commands within LINE and have her calendar automatically updated, while also getting relevant emails filtered and summarized—all inside her chat app?

What This Automation Does

This tailored n8n workflow listens to LINE chat messages and acts as Yuki’s virtual assistant by leveraging AI and integrated APIs. Here’s what happens when the workflow runs:

• Real-time chat input processing: The LINE Webhook node captures messages Yuki sends in LINE instantly.
• Smart message interpretation: AI Agent node with LangChain and OpenAI transforms natural language commands into structured tasks.
• Google Calendar event management: It reads and creates calendar events based on Yuki’s requests, such as “Schedule meeting on Monday 3 PM”.
• Email retrieval from Gmail: Pulls recent emails filtered by dates Yuki mentions in chat and summarizes them for quick review.
• Conversational memory: Window Buffer Memory node maintains chat context for smooth, multi-step interactions.
• Clear responses inside LINE chats: Replies with confirmations, event details, or error messages directly back to Yuki’s LINE account.

With these outcomes, Yuki can reclaim several hours weekly, eliminate manual data entry errors, and enjoy seamless chat-driven scheduling and email monitoring.

Prerequisites ⚙️

• n8n account (cloud or self-hosted) 🔌
• LINE Messaging API webhook setup (access to LINE Developer Console) 💬
• Google Calendar account with OAuth2 credentials 🔑
• Gmail account with OAuth2 credentials 📧
• OpenAI API key configured in n8n for LangChain AI nodes 🔐

Optional: Self-hosting your n8n workflow ensures data privacy and control. Check out Hostinger’s n8n self-hosting guide if interested.

Step-by-Step Guide to Build the Workflow

Step 1: Set up the LINE Receiving Webhook node
Navigate to the n8n Editor UI. Click + Add Node → search for Webhook node → name it “Line Receiving”.
Configure the node with HTTP Method: POST and set the Path to “linechatbotagent”.
You should see a generated webhook URL; copy this to your LINE Developer Console as the webhook URL.
Expected outcome: LINE messages sent to your bot trigger this webhook.
Common mistake: Forgetting to enable the webhook in LINE Developer Console or incorrect webhook path.

Step 2: Add Switch Between Text and Others node
Add a Switch node named “Switch Between Text and Others”.
Configure a rule to check if the incoming LINE message type equals “text”.
If true, direct flow to AI processing, else to error handling.
Outcome: Filters only text messages for AI processing.
Tip: Ensure you map the input data path to $(‘Line Receiving’).item.json.body.events[0].message.type exactly to avoid false negatives.

Step 3: Configure the AI Agent node
Add the LangChain AI Agent node.
Set the “text” parameter to reference the incoming message text: =$json.body.events[0].message.text.
Under options, add a system message such as “You are a helpful assistant. Here is the current date {{ $now }}”.
This node integrates with OpenAI Chat Model, Wikipedia, Gmail Read, and Google Calendar nodes for advanced responses.
Expected: AI understands user intents like querying calendar or emails.
Common mistake: Missing correct integration with linked AI nodes.

Step 4: Set up Google Calendar Read node
Add the Google Calendar Read node.
Configure it to fetch up to 5 events filtered by start and end dates dynamically extracted by AI.
Connect the node’s outputs as AI Agent tools.
Outcome: Enables reading calendar events based on chat instructions.
Tip: Use AI variables correctly in timeMin and timeMax fields for accurate filtering.

Step 5: Set up Google Calendar Create node
Add the Google Calendar Create node.
Configure start and end date/time dynamically from AI output variables (e.g., createStartDate, createEndDate).
Set event summary from AI variable “event_name”.
Connect it to the AI Agent’s tools.
Expected outcome: New events auto-created upon user request.
Note: Google Calendar OAuth2 credentials must be set.
Common mistake: Date/time format errors from AI input.

Step 6: Configure Gmail Read node
Add the Gmail Read node.
Set it to retrieve up to 5 emails filtered by received date extracted by AI.
Connect it as a tool node in AI Agent.
Outcome: Fetches relevant emails for user queries.
Important: Ensure Gmail OAuth2 credentials are correctly set.
Tip: Date format should match Gmail query syntax.

Step 7: Add Window Buffer Memory node
Add LangChain’s Window Buffer Memory node.
Configure sessionKey to userId: =$json.body.events[0].source.userId.
This maintains chat context for multi-step conversations.
Expected result: Conversations stay coherent.
Mistake to avoid: Incorrect session key mapping leading to lost context.

Step 8: Add OpenAI Chat Model node
Add OpenAI Chat Model node.
Leave default settings or add options as needed.
Connect as languageModel input to AI Agent for text generation.
Outcome: High-quality AI responses.
Tip: Use appropriate OpenAI API key in credentials.

Step 9: Add Error Handling from AI Response node (Switch)
Insert a Switch node to check if AI output message exists.
If yes, continue with Text Cleansing and reply; else direct to error reply.
This prevents sending empty messages to users.
Common mistake: Misconfigured condition causing workflow to terminate prematurely.

Step 10: Set up Text Cleansing node
Add a Set node to clean AI response text.
Use transformation logic to strip markdown, newlines, and tags:

=$json.message.content.replaceAll("n","\n").replaceAll("n","").removeMarkdown().removeTags().replaceAll('"',"")

Expected outcome: Clean text reply suitable for LINE messages.
Mistake: Not applying cleansing may cause format errors in LINE.

Step 11: Configure Line Answering nodes (Ordinary and Error Cases)
Use HTTP Request nodes to send replies back to LINE Messaging API.
Set URL to “https://api.line.me/v2/bot/message/reply” with POST method.
In Ordinary Case node, dynamically set message text from cleaned AI output.
In Error Case node, send friendly message like “กรุณาส่งอย่างอื่นเถอะนะเตงอัว” (Please send something else).
Expected: User receives appropriate response quickly.
Common mistake: Incorrect Authorization header or payload format.

Customizations ✏️

• Change default error message: In the “Line Answering (Error Case)” HTTP Request node, edit the JSON body text field to a message better suited to your audience’s language and tone.
• Add support for handling image or sticker messages: Extend the “Switch Between Text and Others” node with additional conditions to process images or other message types with alternative flows.
• Extend AI capabilities with more tools: Connect additional LangChain tools or custom APIs in the AI Agent node to answer more complex queries (e.g., integrate CRM data or weather APIs).
• Customize calendar event details: In “Google Calendar Create” node, add more fields like location, attendees, or description by mapping AI variables accordingly.

Troubleshooting 🔧

• Problem: “401 Unauthorized” error when sending reply to LINE.
  Cause: The LINE bot token in HTTP headers is missing or invalid.
  Solution: Go to the “Line Answering” HTTP Request nodes and confirm the Authorization Bearer token matches your LINE Channel Access Token.
• Problem: AI Agent returns empty or irrelevant answers.
  Cause: Missing or incorrect integration of AI tools or improper prompt system message.
  Solution: Review AI Agent node parameters to ensure AI tools are connected and that the systemMessage is meaningful and up-to-date.
• Problem: Google Calendar events not creating.
  Cause: Date/time extracted is invalid or OAuth2 credentials outdated.
  Solution: Verify the date/time output from AI, ensure proper date formats, and refresh OAuth2 credentials.

Pre-Production Checklist ✅

• Confirm LINE webhook is active and responding in developer console.
• Test sample LINE messages contain text type to pass through the switch node.
• Validate OAuth2 credentials for Google Calendar and Gmail by running test fetches.
• Test AI Agent with sample queries on calendar events and emails.
• Backup your workflow JSON before deployment.

Deployment Guide

Activate your workflow by toggling the active switch in n8n UI.
Monitor the execution logs for errors or failures.
Since this workflow handles live chat commands, keep n8n running reliably with stable internet and server uptime.
Revisit and tweak AI prompts periodically to improve accuracy.
For high usage, consider self-hosting n8n to handle volume and security.

FAQs

• Can I replace Google Calendar with another calendar service?
  Yes, but you’ll need to modify nodes or create custom API calls to suit that calendar’s API.
• Does this consume OpenAI API credits?
  Yes, each AI interaction uses OpenAI tokens that could contribute to cost depending on your usage.
• Is user chat data stored?
  Context is stored temporarily via the Window Buffer Memory node, but persistent storage depends on your n8n setup.
• Can this workflow handle multiple LINE users simultaneously?
  Yes, the sessionKey based on userId keeps conversation context distinct per user.

Conclusion

In building this unique n8n workflow, you’ve automated critical aspects of managing your calendar and email directly from your LINE chat. This integration converts natural language requests into actionable calendar events and filtered email reads, saving hours of manual effort weekly.

By deploying this automation, Yuki—or anyone relying heavily on LINE messaging and Google workspace—will reduce errors and distractions. You could next explore adding voice command support, multi-language translation, or broadening AI-driven task handling with custom APIs.

Keep experimenting, and enjoy your newly freed time and improved productivity!