Route and log incoming emails with GPT-4, Excel 365 and Telegram

Name: Route and log incoming emails with GPT-4, Excel 365 and Telegram
Availability: InStock
Author: Ramdoni

Route and log incoming emails with GPT-4, Excel 365 and Telegram preview

Open on n8n.io

$20/month : Unlimited workflows

2500 executions/month

Try free

THE #1 IN WEB SCRAPING

Scrape any website without limits

Try free

HOSTINGER

Early Deal
DISCOUNT 20%

Self-hosted n8n

Unlimited workflows - from $4.99/mo

Try free

#1 hub for scraping, AI & automation

6000+ actors - $5 credits/mo

Try free

Overview

AI Email Analyzer with Department Routing and Excel Logging (n8n Workflow)

Overview

This workflow automatically analyzes incoming emails using AI, categorizes them, routes them to the appropriate department, logs the results in Excel 365, and sends notifications to the responsible team member via Telegram.

It helps teams automate email triage, prioritize issues, and ensure that important messages are routed to the correct department without manual sorting.

This workflow is designed to be scalable, easy to configure, and adaptable for multiple business use cases such as support management, internal operations, finance monitoring, and technical issue tracking.

Features

AI-powered email analysis
Automatic email categorization
Department routing based on AI classification
Priority detection (High / Medium / Low)
Attachment detection
Action-required detection
Spam filtering
Excel 365 email logging
Department contact lookup from Excel
Telegram notifications to the responsible team member
Modular workflow design for easy extension

How It Works

Email Trigger

The workflow listens for new incoming emails using IMAP, Gmail, or Outlook.

Email Data Extraction

Key email information such as sender, subject, and body content is extracted and prepared for analysis.

Email Body Limiting

The email body is truncated to avoid excessive token usage when sending content to the AI model.

Batch Processing

Emails are processed one by one to prevent rate limits when calling the AI API.

AI Email Analysis

The email is sent to an AI model that analyzes the message and returns structured JSON data including:
summary
category
priority
department
attachment detection
action required
spam detection

Department Lookup

The workflow looks up the responsible department contact using an Excel 365 sheet that maps departments to their respective PIC (Person in Charge).

Email Logging

The analyzed email data is appended to an Excel sheet for tracking and auditing.

Telegram Notification

The responsible department PIC receives a Telegram notification containing the email summary and priority details.

Setup Steps

1. Import the Workflow

Import the workflow JSON file into your n8n instance.

2. Configure Email Trigger

Connect your email account using one of the following: - IMAP - Gmail - Outlook

Set the trigger to detect new incoming emails.

3. Configure OpenAI

Add your OpenAI API credentials in n8n.

The AI model is used to analyze incoming email content and generate structured data such as summary, category, department, and priority.

4. Prepare the Excel File

Create an Excel file with two sheets.

File example: Email_Automation.xlsx

Sheet 1: Email_Log

Timestamp Sender Subject Category Priority Department Action Attachment Spam Summary Required

This sheet stores the processed email records for auditing and tracking.

Sheet 2: Department_Routing

Department PIC_Name Email Telegram_Chat_ID

Support Andi [email protected] 12345678 Sales Rina [email protected] 23456789 Finance Budi [email protected] 34567890 Technical Dodi [email protected] 45678901 Management CEO [email protected] 56789012 General Admin [email protected] 67890123

This sheet maps departments to the responsible team member and their Telegram chat ID.

5. Configure Telegram

Create a Telegram Bot and connect it to n8n.

Ensure the Telegram Chat IDs in the Excel routing sheet are correct so notifications are delivered to the right person.

6. Configure Excel Credentials

Connect your Excel 365 account to n8n and select the Excel file used for logging and department routing.

7. Test the Workflow

Send a test email to your monitored inbox.

The workflow should: 1. Analyze the email using AI 2. Determine category and department 3. Log the result in Excel 4. Send a Telegram notification to the responsible department

Example Output

Example AI output:

{
 "summary": "Customer asking about invoice payment status",
 "category": "Invoice",
 "priority": "Medium",
 "department": "Finance",
 "has_attachment": true,
 "action_required": true,
 "spam": false
}

Example Telegram notification:

📩 New Email Routed

Department: Finance
Priority: Medium

From: [email protected]
Subject: Invoice payment status

Summary:
Customer asking about invoice payment status

Action Required: Yes
Attachment: Yes

Requirements

n8n
OpenAI API key
Email account (IMAP / Gmail / Outlook)
Microsoft Excel 365
Telegram Bot

Customization Ideas

You can extend this workflow by adding: - Jira ticket creation for technical issues - Slack or Microsoft Teams notifications - CRM integration for sales leads - sentiment analysis for customer emails - automated support ticket creation - SLA monitoring and escalation

Notes

This workflow demonstrates how AI can automate email intake, classification, and operational routing for teams.

It can be adapted for customer support, internal operations, finance monitoring, and technical issue tracking.

Ramdoni

6 workflows

Complexity

advanced

Published 11 Mar 2026

Likes 0

View on n8n.io Download Workflow

Install path: /data/workflows/13989/13989.json

Share Your Workflow

Have a useful automation to share? Publish it and help the community.

Submit Your Template How to Submit

Related Workflows

Process website contact form submissions with SharePoint and Gmail

How it works This workflow fires whenever a visitor submits a contact form on your website. 1. Webhook receives the POST payload (Name, Email, Phone, Message, Date, Time) from your website form. 2. Log Submission to SharePoint writes each field into a SharePoint list for permanent record-keeping. 3. Build Branded Email HTML — Code node constructs a fully branded HTML notification email (customize colors and logo to match your brand). 4. Send Email Notification delivers the notification via Gmail to your designated recipient address, with Reply-To automatically set to the submitter's email so you can respond in one click.

View

Triage customer complaint emails from Gmail with GPT-4o-mini, Slack and Sheets

### Description Activate this workflow once and it monitors your Gmail inbox every minute automatically. Every incoming email is scanned for complaint keywords — and if a complaint is detected, GPT-4o-mini scores its urgency, identifies the right department, writes a one-line summary, and drafts a ready-to-send reply. The complaint is logged to Google Sheets and a formatted alert is posted to the correct Slack channel instantly. Built for support teams, customer success managers, and small businesses who can't afford to miss or delay a complaint. --- ## What This Workflow Does - **Automatic inbox monitoring** — Checks your Gmail every minute so no complaint sits unread while your team is busy with other work - **Keyword-based complaint detection** — Scans subject lines and email bodies for 12 complaint signals and silently ignores everything else - **AI urgency scoring** — Rates every complaint from 1 to 10 so your team knows at a glance what needs attention first - **Department routing** — Classifies each complaint as Billing, Technical, or General and sends the alert to the right Slack channel automatically - **Reply draft generation** — GPT-4o-mini writes a professional, empathetic reply your agent can send immediately — no drafting from scratch - **Permanent complaint log** — Appends a 12-column record to Google Sheets for every complaint, giving you a searchable history for reporting and pattern analysis - **Clean non-complaint exit** — Emails that are not complaints pass through silently without triggering any logging or alerts --- ## Setup Requirements ### Tools Needed - n8n instance (self-hosted or cloud) - Gmail account (the inbox you want monitored) - OpenAI account with GPT-4o-mini API access - Slack workspace with OAuth2 app configured - Google Sheets (one sheet with a tab named **Complaint Log**) ### Credentials Required - Gmail OAuth2 - OpenAI API key - Slack OAuth2 - Google Sheets OAuth2 **Estimated Setup Time: 15–20 minutes** --- ## Step-by-Step Setup 1. **Import the workflow** — Open n8n → Workflows → Import from JSON → paste the workflow JSON → click Import 2. **Connect Gmail** — Open node **1. Gmail — Inbox Monitor** → click the credential dropdown → select OAuth2 → sign in with the Gmail account you want monitored → authorize access 3. **Fill in your config values** — Open node **2. Set — Config Values** → replace all seven placeholders: | Field | What to put here | |---|---| | `YOUR COMPANY NAME` | Your business name (used in the AI reply) | | `YOUR_GOOGLE_SHEET_ID` | The ID from your Google Sheet URL (the string between `/d/` and `/edit`) | | `Complaint Log` | Leave as-is, or match your sheet tab name exactly | | `#billing-support` | Your Slack channel for billing complaints | | `#tech-support` | Your Slack channel for technical complaints | | `#customer-support` | Your Slack channel for general complaints | | `[email protected]` | Your support team's reply-from email | 4. **Connect OpenAI** — Open node **6. OpenAI — GPT-4o-mini Model** → click the credential dropdown → add your OpenAI API key → test the connection 5. **Connect Google Sheets** — Open node **9. Google Sheets — Log Complaint** → click the credential dropdown → connect your Google account via OAuth2 → make sure your sheet has a tab named exactly **Complaint Log** with these 12 column headers in row 1: `Email ID`, `Received Date`, `Sender Name`, `Sender Email`, `Subject`, `One Line Summary`, `Department`, `Urgency Score`, `Urgency Level`, `Suggested Reply`, `Slack Channel Alerted`, `Logged At` 6. **Connect Slack** — Open node **10. Slack — Send Department Alert** → click the credential dropdown → connect your Slack workspace via OAuth2 → invite the n8n bot to all three complaint channels in Slack (`/invite @n8n` in each channel) 7. **Activate the workflow** — Toggle the workflow to Active — it will begin polling Gmail every minute immediately --- ## How It Works (Step by Step) **Step 1 — Gmail: Inbox Monitor** This step checks your Gmail inbox every 60 seconds for new emails. Every new message that arrives is passed to the next step for processing. Nothing is read, flagged, or deleted — it is read-only access. **Step 2 — Set: Config Values** Your company name, Google Sheet ID, sheet tab name, three Slack channel names, and support email are stored here as named variables. You set these once and every other step in the workflow uses them automatically. **Step 3 — Code: Extract Email Fields** This step pulls the sender name, sender email, subject line, and email body out of the raw Gmail data. The email body is trimmed to 3,000 characters to keep AI processing fast and cost-efficient. A timestamp is also captured for the log. **Step 4 — IF: Is This a Complaint?** This is the filter step. It checks whether the subject or body contains any of 12 complaint signals: words like "complaint", "refund", "not working", "unacceptable", "very unhappy", "demand refund", and others. If a match is found (YES path), the email moves forward to AI triage. If no match is found (NO path), the email flows to **11. NoOp — Not a Complaint** and the workflow stops silently — no logging, no alert. **Step 5 — AI Agent: Triage Complaint** GPT-4o-mini reads the sender name, subject, and email body. It returns exactly four pieces of data: an urgency score from 1 to 10, the correct department (Billing, Technical, or General), a one-line summary of the complaint, and a ready-to-send reply draft under 80 words addressed to the customer by name. **Step 6 — OpenAI: GPT-4o-mini Model** This is the language model powering the triage step. It runs at temperature 0.3 for consistent, structured responses and is capped at 600 tokens to keep costs low per complaint. **Step 7 — Parser: Structured Triage Output** This step enforces the exact data structure GPT-4o-mini must return. It validates that urgency score is a number, department is one of the three valid options, and both the summary and reply fields are present. This prevents malformed AI output from reaching your sheet or Slack. **Step 8 — Code: Combine Triage Data** All AI results are merged with the original email data here. This step also converts the numeric urgency score into a human-readable label: 1–3 = Low, 4–5 = Medium, 6–7 = High, 8–10 = CRITICAL. It also selects the correct Slack channel based on the department GPT-4o-mini assigned. **Step 9 — Google Sheets: Log Complaint** A new row is appended to your Complaint Log sheet with all 12 fields: email ID, received date, sender details, subject, AI summary, department, urgency score, urgency label, suggested reply, which Slack channel was alerted, and the time it was logged. **Step 10 — Slack: Send Department Alert** A formatted Slack message is posted to the correct channel — `#billing-support`, `#tech-support`, or `#customer-support` — depending on the department. The message shows the urgency label, score, sender details, subject, AI summary, and the suggested reply text ready for the agent to copy and send. The final result: your team sees a structured complaint alert in Slack within seconds of the email arriving, with the reply already written. --- ## Key Features ✅ **Runs without any manual trigger** — Gmail polling fires every minute automatically once the workflow is active ✅ **Zero noise for non-complaints** — Regular emails pass through silently with no logging, no alerts, and no wasted API calls ✅ **CRITICAL flag for urgent cases** — Any complaint scoring 8 or above is labelled CRITICAL in Slack so high-risk customers get immediate attention ✅ **Ready-to-send reply included** — Every Slack alert contains a drafted reply addressed to the customer by name — agents copy, review, and send ✅ **Structured AI output enforced** — A schema parser ensures GPT-4o-mini always returns the correct fields in the correct format, preventing broken logs ✅ **Three-channel Slack routing** — Billing, Technical, and General complaints each go to their own channel automatically — no manual sorting ✅ **12-column complaint history** — Every complaint is permanently recorded in Google Sheets for weekly reporting, pattern spotting, and team reviews ✅ **Token-efficient processing** — Email body is capped at 3,000 characters and AI is capped at 600 tokens — keeping costs predictable at scale --- ## Customisation Options **Add more complaint keywords** — In node **4. IF — Is This a Complaint?**, add more trigger phrases to the condition (e.g. `very frustrated`, `cancel my account`, `escalate`) to catch complaints your current list misses. **Adjust urgency thresholds** — In node **8. Code — Combine Triage Data**, change the score cutoffs for Low, Medium, High, and CRITICAL to match your team's definition of urgency (e.g. raise CRITICAL from 8 to 9 for a stricter threshold). **Add a Gmail label to processed complaints** — After node **10. Slack — Send Department Alert**, add a Gmail node set to "Add Label" to tag every processed complaint in your inbox (e.g. "Triaged") so agents know which emails the workflow has already handled. **Send a copy to email** — After node **8. Code — Combine Triage Data**, add a Gmail Send node to also email the suggested reply and triage details to your support team inbox as a backup alongside Slack. **Weekly summary report** — Add a Schedule trigger that runs every Monday morning, reads the Google Sheet via a Sheets node, counts complaints by department and urgency level, and posts a summary to a `#support-weekly` Slack channel. **Route CRITICAL complaints to a separate Slack channel** — In node **8. Code — Combine Triage Data**, add a condition: if `urgencyLabel === 'CRITICAL'`, override `slackChannel` with a dedicated `#escalations` channel so your most urgent cases never get buried. --- ## Troubleshooting **Gmail not triggering the workflow:** - Make sure the workflow is toggled to Active — inactive workflows do not poll - Check that the Gmail OAuth2 credential in node **1. Gmail — Inbox Monitor** is connected and not expired — re-authorize if needed - Send a test email to the monitored inbox and wait up to 60 seconds for the next poll cycle **OpenAI credential not working:** - Confirm the API key is connected in node **6. OpenAI — GPT-4o-mini Model**, not in a different step - Check that your OpenAI account has available credits - Verify the key has access to gpt-4o-mini — restricted keys may block this model **Google Sheets not logging rows:** - Confirm the Sheet ID in node **2. Set — Config Values** matches the ID in your Google Sheet URL exactly - Make sure the tab is named **Complaint Log** — the name must match `sheetName` in Config Values exactly, including capitalization - Check that the Google Sheets OAuth2 credential in node **9. Google Sheets — Log Complaint** is connected and authorized **Slack alerts not arriving:** - Confirm the Slack OAuth2 credential in node **10. Slack — Send Department Alert** is connected - Make sure the n8n bot has been invited to all three channels — type `/invite @n8n` in each channel in Slack - Check that the channel names in node **2. Set — Config Values** include the `#` prefix and match exactly **AI returning wrong department or malformed output:** - The structured output parser in node **7. Parser — Structured Triage Output** enforces the schema — if GPT returns an unexpected format, check the execution log of node **5. AI Agent — Triage Complaint** for the raw output - If department values are inconsistent, the prompt in node **5** explicitly requires exactly `Billing`, `Technical`, or `General` — any deviation means the AI call failed and you should re-run --- ## Support Need help setting this up or want a custom version built for your team or agency? 📧 Email: [[email protected]]([email protected]) 🌐 Website: [https://www.incrementors.com/contact-us/](https://www.incrementors.com/contact-us/)

View

Triage customer support emails and draft Gmail replies with IONOS AI Model Hub

## AI Customer Support Triage — Auto-Draft Replies with IONOS AI Model Hub This n8n template shows you how to automate customer support email triage with a sovereign AI. By combining Gmail for inbox monitoring and the IONOS AI Model Hub, every customer email is classified and replied to by AI. ## Use cases - **Support inbox management:** Automatically sort incoming emails into Spam, Sales Lead, Tech Support, FAQ, Billing, or Other — so your team knows exactly what needs attention and in what order. - **Draft replies in seconds:** Instead of writing responses from scratch, your agent opens Gmail Drafts to find a ready-to-send reply already written in the customer's language — just review and hit Send. - **Multi-inbox coverage:** Monitor both your info@ and support@ addresses or more in a single workflow, with a filter that ensures only legitimate business emails reach the AI. ## How it works A Gmail Trigger polls your inbox every minute for new unread messages. A Filter node immediately discards anything not addressed to your support@ or info@ address, so irrelevant mail never reaches the AI. The **Prepare Email Data** code node strips HTML, removes quoted reply chains, and extracts the clean fields — sender, subject, date, recipient, and body — with no prompt logic mixed in. The **Basic LLM Chain** node then takes over: it sends the cleaned email to the IONOS AI Model Hub (Mistral Small 24B) with a structured prompt that instructs the model to return a single JSON object containing the category, priority, sentiment, detected language, a one-sentence summary, and a full draft reply written in the same language as the customer's email. A **Parse AI Response** code node extracts and validates the JSON, with a graceful fallback if the model output is malformed. A Switch node routes Spam emails to a silent mark-as-read action, while all other categories trigger the **Gmail Create Draft** node — which saves the AI-written reply directly into your Drafts folder, threaded to the original conversation. Your agent reviews it and clicks Send. ## Good to know **Filtering:** The Filter node uses a partial match on `support@` and `info@`. For a stricter match, update the values to your full addresses (e.g. `[email protected]`). **Model selection:** The IONOS AI Model Hub offers several LLMs including Mistral Small 24B, and Nemo. For email triage and reply drafting, Mistral Small 24B the best balance of reasoning quality and response speed. **Language support:** The prompt instructs the model to detect the email language and reply in kind — no additional configuration needed for multilingual inboxes. ## How to set it up 1. Connect your Gmail OAuth2 credentials to the Gmail Trigger, Filter, and both Gmail action nodes. 2. Update the Filter node values with your actual support addresses. 3. Add your IONOS Cloud API token to the **IONOS Cloud Chat Model** node credentials. 4. Customize the prompt in the **AI Triage & Draft Reply** node — adjust categories, tone, or signature as needed. 5. Activate the workflow and send a test email to verify the draft appears in Gmail. ## Requirements - Gmail account (OAuth2) for inbox monitoring and draft creation - IONOS Cloud account to access the AI Model Hub

View

Need Custom Automation?

Get help designing a custom n8n workflow that connects your stack and fits your process.

Route and log incoming emails with GPT-4, Excel 365 and Telegram

Workflow preview

Overview

AI Email Analyzer with Department Routing and Excel Logging (n8n Workflow)

Overview

Features

How It Works

Setup Steps

1. Import the Workflow

2. Configure Email Trigger

3. Configure OpenAI

4. Prepare the Excel File

Sheet 1: Email_Log

Sheet 2: Department_Routing

5. Configure Telegram

6. Configure Excel Credentials

7. Test the Workflow

Example Output

Requirements

Customization Ideas

Notes