Mahar Automation

One shop. Everything connected.

Multimodal customer service agent handling Arabic dialect voice notes and text orders.

Retail automation and AI. Built with Facebook Messenger API / Whisper STT / Human Override / n8n / AI Agent

The messaging agent workflow in n8n: an incoming webhook, voice-note transcription, a text classifier, then an AI agent that answers from the shop Sheets data. — Actual n8n workflow

Behind the build.

The problem

A seafood business owner received continuous voice notes and dialect inquiries on Facebook Messenger that could not be processed by standard rigid chatbots.

What I built

An n8n conversational agent powered by Whisper speech-to-text and chat models tuned for Egyptian Arabic. Integrates human agent takeover detection that immediately pauses automated replies when staff intervenes, and links customer orders to inventory.

The result

Zero customer wait time for common price and stock inquiries with automated human agent override protection.

The Mahar Mart agent handling a customer in Egyptian Arabic: it greets them by name, lists herring, marinade and shrimp, and sends a photo of the packed product. — أهلاً بك يا أستاذ. The agent answers in the dialect customers actually use.

Need something like this?

Tell me where the work gets stuck. Let’s figure out what we can build.

Let’s talk