Pick n Pay

Pick n Pay launched 'Penny', an AI-powered shopping assistant leveraging Google's Gemini AI, enabling customers to order via voice, text, or photos with recipe suggestions and personalized recommendations.

Leadership read: Pick n Pay has committed itself to an AI-native customer interface as a primary ordering channel, not a feature bolted onto an existing app. Accepting voice, image, and text inputs simultaneously requires a fundamentally different data architecture than a conventional search-and-basket model: intent recognition, multimodal ingestion, and real-time inventory mapping all have to resolve together at the moment of query. By embedding meal planning and personalized recommendations inside that same flow, the company has also made the assistant responsible for basket value, not just order completion. That is an ownership of commercial outcome that a search box never had. The related signals here are thematically adjacent rather than sector-matched. Of the 12 product launches we tracked in this period, the closer analogues are Starling's client-facing AI tooling and Raft's human-agent collaboration platform, both pointing to the same structural pattern: incumbents in established sectors embedding persistent AI agents into workflows previously handled by human navigation or manual input. The category is broad, but the directional read is consistent: AI-native customer interfaces are moving from pilot status to primary-channel investment across retail, fintech, and logistics simultaneously. Across companies making this transition, the pattern surfaces concentrated demand for product leadership at the seam between machine-learning infrastructure and customer experience, commercial operators who can manage performance against basket metrics through an AI layer rather than direct merchandising controls, and data engineering capability sufficient to sustain personalization at scale in markets where behavioral datasets are still being built.

curated · 2026-07-02 · context →