Pick n Pay Launches Penny AI Assistant on asap! Using Google Gemini

The retailer has unveiled Penny, an AI-powered shopping assistant integrated into its on-demand grocery platform, asap!, marking one of the most significant technology upgrades to the service since its launch.

Built using Google's Gemini large language models, Penny enables customers to build shopping baskets using natural conversation through voice, text or even photographs.

The feature, which begins rolling out in July, positions Pick n Pay firmly in the next phase of South Africa's fast-growing e-commerce battle, where retailers are increasingly using artificial intelligence to differentiate customer experiences rather than competing solely on delivery speed.

A shift from search to conversation

Unlike traditional online grocery shopping, where customers browse categories or search for individual products, Penny allows shoppers to describe what they need in everyday language.

Customers can ask the assistant to recommend meals, generate shopping lists, suggest alternatives, build baskets within a budget or even upload handwritten shopping lists and photographs of products for recognition.

Need ingredients for a pasta dinner? Penny provides recipe suggestions before automatically assembling the required products into a shopping basket. Running low on groceries? Customers can photograph the contents of their fridge and receive meal recommendations based on what they already have.

"We're moving from search-and-scroll shopping to conversation-led shopping that makes buying groceries faster, smarter and far more intuitive," said Enrico Ferigolli, retail executive for omnichannel at Pick n Pay at the unveiling.

"For years, the focus has been on faster delivery. The next disruption is removing the effort from shopping itself."

The multimodal assistant can process voice notes, typed requests and images, while recognising multiple languages, including several South African languages, making the experience significantly more accessible for local consumers.

Image supplied

The next chapter in retail innovation

The launch builds on a major overhaul of the asap! platform completed in 2025, when Pick n Pay rebuilt its online ecosystem to integrate grocery shopping, Smart Shopper rewards and value-added services into a single platform.

Since then, the retailer has expanded the service to include Pick n Pay Clothing and continued investing heavily in digital capabilities.

Those investments are already translating into strong growth.

According to Pick n Pay's latest financial results, its online business recorded turnover growth of 32.7% on a like-for-like basis during the 2026 financial year, while its on-demand delivery services, including asap! and Pick n Pay groceries on Mr D, increased by 37.6% year-on-year.

Ferigolli described the platform rebuild as the foundation that made AI integration possible.

"This is only the beginning of what AI can unlock for everyday grocery shopping, and Penny is the first step in that journey."

Competing in South Africa's AI grocery race

The introduction of Penny comes just months after rival Checkers Sixty60 introduced its own AI assistant, Pixie, signalling a new competitive front in South Africa's grocery retail market.

While both assistants leverage artificial intelligence, their approaches differ significantly.

Pick n Pay has positioned Penny as a conversational shopping companion capable of understanding context and responding naturally to customer requests.

Shoprite's Pixie, by comparison, currently focuses more heavily on predictive replenishment and personalised product recommendations based on previous shopping behaviour.

Ferigolli believes the conversational interface creates a more intuitive shopping experience by allowing customers to interact with the platform as they would with another person.

"By understanding what customers are cooking, buying and planning, Penny becomes more than a shopping app. It becomes a genuine shopping companion."

Powered by Google's Gemini AI

Penny is built on Google's Gemini foundation models, with Pick n Pay selecting the platform after evaluating several leading large language models.

Rather than accessing Pick n Pay's systems directly, Gemini works alongside dedicated application programming interfaces (APIs) connected to the retailer's product catalogue, Smart Shopper loyalty programme, order history and curated retail content.

According to Google South Africa country director Kabelo Makwane, the partnership reflects a broader evolution in artificial intelligence.

"AI is moving beyond simply answering questions to helping people get things done," said Makwane.

"The power of multimodal AI is that people don't have to adapt to technology — the technology adapts to them."

Building a smarter shopping experience

Beyond simplifying grocery shopping, Penny has been designed to become increasingly personalised over time.

As customers continue using the assistant, it learns shopping preferences including favourite brands, purchasing habits, price sensitivity and previous orders, enabling increasingly tailored recommendations.

Future developments are expected to include deeper personalisation, expanded recommendation capabilities and additional AI-powered services as Pick n Pay continues investing in the platform.

For now, customers remain in control of checkout, payment and delivery scheduling, with Penny focused on removing friction from product discovery and basket building.

Convenience becomes the new competitive advantage

Artificial intelligence is rapidly becoming the next battleground in South Africa's grocery sector as retailers compete to deliver more seamless customer experiences.

For Pick n Pay, Penny represents more than another digital feature.

It reflects a broader shift in retail strategy — one where convenience is measured not only by how quickly groceries arrive, but by how little effort customers expend before clicking "checkout".

As South African consumers increasingly embrace conversational AI across everyday services, grocery shopping may soon become another activity driven less by searching and more by simply asking.