How I Built MzansiShopper, An AI-Powered Retail Assistant for Personalized Shopping

Introduction
Imagine walking into a store and the assistant knows exactly what you’re looking for, even speaks your language. That’s the shopping experience I wanted to bring online in South Africa.

The Problem

• There is limited personalization in retail in South Africa.

• 11 official languages hence creating accessibility barriers.

• Consumers often feel excluded from digital tools.

I watched my mom struggle to shop online because everything was in English.

MzansiShopper is my solution to make retail more personalized and inclusive using AI.

South Africa's diverse languages and shopping needs inspired me to build a smart assistant that understands and recommends products in real time powered entirely by AWS services and deployed with Q- CLI for automation.

The Solution

My goal was simple, to build an AI assistant that delivers personalized product recommendations in multiple languages, helping customers find exactly what they need, while supporting digital inclusion in South Africa's retail sector.

But then came the challenge of building it. That's when I remembered Q CLI. Instead of starting from scratch, I leveraged Amazon Q for CLI, interacting with it right from my terminal to generate python code and bring MzansiShopper to life.

Stack Overview

MzansiShopper leverages AWS services:

• Lex powers the conversational chat interface, understanding natural language.

• Personalize generates real-time tailored product suggestions.

• Translate enables multilinguals support, breaking language barriers.

• DynamoDB securely stores user data and shopping history.
  And I used Q- CLI for streamlined deployment and management of all these resources.

The process

Step by step Q-CLI created and built the project.
But it didn’t stop there! It also walked me through the installation process and how to run the app successfully.
Then… BOOM! It worked on the first try—no debugging, no errors, just pure Amazon Q magic. 

The End Product

Deployment Process

• Automating setup with CLI.

• Handling data for recommendations.

• Integrating Translate for multilingual support.

Lessons learned

• Challenge: Managing multiple languages.

• Win: CLI automation saved tons of time.

• Surprise: How quickly AWS Personalize started delivering accurate results.

The Impact

• MzansiShopper has enables digital inclusion.

• Small retailers can now compete with bigger players.

• Shoppers across languages feel seen and understood.

If you’re building with AWS, these tools are not for Silicon Valley startups, they’re for solving problems right here in Africa.

With the right tools, we can build solutions that are inclusive, scalable, and impactful.

MzansiShopper is a step towards making every South African feel at home in the digital economy.
Building with Amazon Q-CLI was an absolute blast! . It’s truly a game-changer fast, intuitive, and super helpful.
Supported environment for Amazon Q for command line
Install Amazon Q for command line in your own environment
Live Mzanzi App
Blog About Q-CLI