An AI-powered system that automatically processes customer calls, analyzes their sentiment, and provides statistics through an AI chat interface.

1. Context

Customer support and sales teams receive valuable feedback through phone conversations. However, as call volumes grow, it becomes increasingly difficult to analyze recordings, track customer sentiment, and identify recurring issues.

This proof of concept demonstrates how AI can turn customer calls into structured, accessible business insights.

2. Problem

Valuable customer feedback gets lost in hours of call recordings.

When calls are reviewed manually, businesses face three key challenges:

• Time-consuming processes: Employees spend hours listening to recordings, creating transcripts, and preparing reports.
• Missed customer issues: Negative experiences and recurring complaints can go unnoticed.
• Slow decision-making: Managers cannot quickly access the statistics they need without manually filtering and analyzing data.

The goal: Make customer feedback easier to understand, reduce repetitive work, and provide clear call statistics whenever they are needed.

3. Solution

The solution consists of three interconnected workflows built in n8n.

System architecture consisting of three interconnected workflows.

1. Call Transcription Processing

The first workflow handles the main call-processing tasks:

• Retrieves the call recording that needs to be processed.