AI-Powered Customer Feedback Automation (Zapier + GPT-4o Mini)

Overview

This project demonstrates a low-code automation pipeline built using Zapier to intelligently process and route customer feedback.

The system collects feedback via Google Forms, applies AI-based sentiment analysis using GPT-4o Mini, and dynamically routes responses based on sentiment score — enabling automated appreciation for positive feedback and real-time internal escalation for negative feedback.

---

Problem Statement

Manual review of customer feedback can be slow and inconsistent. Positive feedback often goes unacknowledged, while negative feedback may not reach the right team quickly.

This workflow automates:

• Feedback intake
• Sentiment classification
• Customer acknowledgment
• Internal escalation

---

Architecture Overview

1️⃣ Trigger – Google Forms

• Customer submits feedback form
• Fields captured:
  • Full Name
  • Email Address
  • Feedback Text

---

2️⃣ Data Formatting – Formatter by Zapier

• Extracts First Name from Full Name
• Prepares clean inputs for downstream personalization

---

3️⃣ AI Sentiment Analysis – AI by Zapier (GPT-4o Mini)

• Model: OpenAI GPT-4o Mini
• Task:
  • Analyze sentiment of feedback text
  • Return structured Boolean score:
    • 1 → Positive sentiment
    • 0 → Negative sentiment

Example prompt logic:

• Classify sentiment
• Return only 1 or 0
• No additional text output

---

4️⃣ Conditional Logic – Zapier Paths

Path A (Score = 1 → Positive Feedback)

• Send personalized thank-you email
• Email includes:
  • Customer first name
  • AI-generated feedback summary
  • Appreciation message

Path B (Score = 0 → Negative Feedback)

• Send WhatsApp notification to internal employee
• Includes:
  • Customer name
  • Contact details
  • Feedback content
• Enables proactive service recovery

---

Workflow Diagram (Logical Flow)

Google Form Submission
→ Formatter (Extract First Name)
→ AI Sentiment Analysis (Boolean 1/0)
→ Paths (Conditional Routing)
→ Positive → Send Thank You Email
→ Negative → Notify Employee via WhatsApp

---

Key Features

• AI-driven sentiment scoring
• Automated customer acknowledgment
• Real-time internal escalation
• Low-code implementation
• Structured Boolean classification logic
• Dynamic personalization

---

Tools Used

• Zapier
• Google Forms
• Formatter by Zapier
• AI by Zapier (OpenAI GPT-4o Mini)
• Email Integration
• WhatsApp Integration

---

Potential Enhancements

• Multi-class sentiment scoring (Positive / Neutral / Negative)
• Severity-based escalation tiers
• Slack integration for internal notifications
• Dashboard for feedback analytics
• Storing structured results in Airtable or Google Sheets

---

Impact

This automation reduces manual triage effort, ensures timely customer acknowledgment, and improves operational responsiveness to negative feedback — demonstrating how AI can be embedded into everyday business workflows using low-code tools.