Project Overview

A fully automated AI‑powered audio processing system built with n8n and integrated with a custom website developed using React, TypeScript, and Material UI. The website serves as the user-facing interface, sending audio requests to n8n via webhooks and displaying structured responses generated by the AI agent.

This project combines frontend development, automation engineering, API orchestration, AI reasoning, and real-time communication between the website and backend workflows.

Context — Business, Industry, Current Process

The company collects customer feedback through recorded phone calls. Managers manually review each audio file and store the results in Google Sheets. A second Google Sheet is used as an inbox: managers paste links to new MP3 files every time a new review is needed.

The goal was to automate:

• transcription of audio feedback
• AI‑based analysis
• statistics generation
• updating Google Sheets
• returning structured results to both the internal chat and the custom website

This system needed to work continuously, reliably, and without human intervention.

Problem — What Was Not Working

Before automation:

• Managers spent hours manually listening to calls.
• Feedback analysis was inconsistent and subjective.
• Google Sheets were updated manually, often with delays.
• No unified statistics or automated reporting existed.