Automated Customer Feedback Triage with Gemini, Jotform, Google Sheets & Gmail Replies

Who’s it for

Teams that collect product feedback with Jotform and want automated triage: instant alerts for urgent issues, knowledge-based replies to questions, and a clean backlog for suggestions.

How it works / What it does

New Jotform submissions are classified by Feedback Type (comments, questions, suggestions) and analyzed for sentiment. Angry or negative messages trigger a Telegram alert to your support channel. Questions are answered automatically using an agent that consults your internal Google Sheet “Q&A” database, then replies via Gmail with a styled HTML email. Suggestions are summarized, sent to a team chat, and appended to a Google Sheets backlog with name, email, summary, and date. All comments are logged to a structured sheet for reporting.

How to set up

Connect your Jotform form and map fields (Name, E-mail, feedback text).

Add credentials for Google Sheets, Gmail, Telegram, and LLM.

Point the agent to your “Q&A” sheet/tab.

Test each branch (comments, questions, suggestions) with sample entries.

Requirements

Jotform account and live form

n8n (self-hosted recommended if using community/LLM nodes)

Google Sheets, Gmail, Telegram bot, and LLM API keys

How to customize the workflow

Adjust sentiment thresholds or add new categories in the Switch node.

Edit the Gmail HTML template and subject line.

Change backlog columns or target sheets/tabs.

Add rate-limits, SLA timers, or different alert channels.

Best practice: rename all nodes, add a yellow Sticky Note with this description and per-step notes, and group user-config variables in a Set node. Remove any hard-coded IDs before sharing

Knowledge base table

Suggestions database

Sentiment database

0
Downloads
2
Views
8.26
Quality Score
intermediate
Complexity
Author:Rully Saputra(View Original →)
Created:10/17/2025
Updated:11/17/2025

🔒 Please log in to import templates to n8n and favorite templates

Workflow Visualization

Loading...

Preparing workflow renderer

Comments (0)

Login to post comments