AI Repair Shop Assistant: Full Customer Journey on WhatsApp (n8n + AI Agents)

I built an AI system that runs a repair company's entire customer process automatically, over WhatsApp.

The problem

A device repair business has to handle many customers at once. Taking in details, sending quotes, chasing approvals, tracking parts, telling people when to pick up, and asking for feedback. Doing all of this by hand is slow and things get missed.

What I built

One large n8n system with a team of AI agents, each doing one job.

Takes in new customers through a webhook and creates their record

Generates a quote PDF and sends it to the customer

Reads the customer's WhatsApp reply and understands what they want (approval, question, complaint, or call request)

Replies in the customer's own language, German, English, or others

Reminds staff when a device is waiting for parts

Creates a pickup code and tells the customer when their device is ready

Waits 7 days after a repair, then asks the customer for feedback

Flags negative feedback as a complaint so the team can act fast

The smart part

Instead of one big confused bot, I used many small AI agents that each handle one clear task. This makes the system reliable and easy to fix. An intent classifier reads every message first and sends it to the right agent.

Tech used

n8n, Google Gemini, WhatsApp (Evolution API), Google Sheets, PDF Generator API, Gmail.

Result

A hands-off assistant that manages the full customer journey from first contact to feedback, so the team can focus on repairs instead of messages.💯