The Future of Shopping: Everything you Need to Know About Walmart Sparky

On June 6, 2025, Walmart launched Sparky, an AI shopping assistant that's now live in their mobile app. This isn't just another search improvement—Sparky represents Walmart's entry into agentic AI. In this article, we'll explain what this means for Walmart sellers, what this means for e-commerce, how technology like Sparky works, and how to optimize your listings for it.

1. What Is Walmart Sparky?

Walmart Sparky is an AI assistant accessible through a smiley face button in the Walmart app. Ask Sparky a question, and it will derive your intent and context to provide a carousel of items along with a description on why these items fit your specific situation.

For example, ask "what's the best laptop for an art student" and Sparky recognizes you need graphics capabilities, adequate storage, and student-friendly pricing. It can handle complex requests like "what sports teams are playing tonight and help me find the right jersey" or provide weather-based outfit recommendations—all in one conversation.

2. Why Walmart Built Sparky: What This Means for E-commerce

Walmart's decision to build Sparky was a response to concrete market signals from their own customer research. Walmart's 2025 Retail Rewired Report revealed that 27% of their shoppers now trust AI recommendations more than influencer endorsements, and 47% would trust AI to purchase household essentials within a budget. The data also showed that 69% of Walmart customers consider shopping speed a key factor in retailer choice, while traditional search methods were creating friction.

This research directly informed Walmart's strategic investment in AI capabilities, and represents a market shift that will force other retailers to make similar investments or risk losing customers to more intelligent shopping experiences. Building features like Sparky and Rufus will become a necessity. To drive this point home, other research shows: Over 58% of consumers have replaced traditional search engines with Gen AI tools for product/service recommendations, and two-thirds of Gen Z and millennials expect hyper-personalized Gen AI powered content and product recommendations.

3. How Walmart Built Sparky's AI Technology

Walmart built Sparky using a hybrid approach that combines third-party AI services with their own proprietary technology. The company has developed internal "Wallaby" AI models trained on retail data, though these aren't fully deployed yet.

Central to Sparky's recommendation accuracy is Walmart's research on "Relevance Filtering for Embedding-based Retrieval," which introduces a 'Cosine Adapter' that maps raw AI similarity scores into interpretable relevance scores, significantly increasing precision in product recommendations. Building on this foundation, their "Enhancing Relevance of Embedding-based Retrieval at Walmart," research introduces Relevance Reward Models that remove noise from training data and handle customer typos, improving the overall quality of Sparky's responses. Of similar importance is Walmart's "MaxMin-RLHF: Equitable Alignment of Large Language Models with Diverse Human Preferences" research, which ensures Sparky serves diverse customer preferences fairly rather than just catering to majority viewpoints.

All of this is built on Walmart's "Element" machine learning platform—a scalable infrastructure that enables rapid development and deployment of AI models across multiple clouds. Element reduces the time to operationalize models from weeks to under an hour, allowing Walmart to quickly iterate and improve Sparky's capabilities.

Of course, the majority of their research, like their solution, is proprietary, but through the research they published throughout 2024, some of the logic can be parsed - and in many ways it surprisingly similar to Amazon Cosmo and Rufus.

4. How Does Walmart Sparky Compare to Amazon Cosmo and Rufus?

Amazon has two main AI initiatives: COSMO (Common Sense Knowledge Generation system) and Rufus (customer-facing chatbot). COSMO uses large language models to build knowledge graphs that understand customer intent—like recognizing that "winter clothes" means garments that provide warmth. Rufus answers product questions using existing data.

Despite Walmart's positioning of Sparky as different, the underlying approaches are strikingly similar. Like COSMO, Sparky focuses on understanding customer intent beyond literal search terms, using large language models to bridge the gap between what customers type and what they actually need.

The seller optimization strategies are remarkably similar across all three systems. Whether optimizing for COSMO's knowledge graphs, Rufus's responses, or Sparky's recommendations, sellers need rich product descriptions that communicate use cases and contexts, helping AI understand when and why customers need specific products. (Something our platform helps check for and generate).

The key differentiator is execution: while COSMO works behind the scenes and Rufus provides Q&A responses, Sparky acts as a conversational shopping companion that guides complete purchase decisions. For sellers, this means optimization strategies developed for Amazon's COSMO algorithm translate well to Sparky, but are not entirely the same.

5. Where can Customers Find Walmart Sparky?

Walmart Sparky is currently only available to US customers in the Walmart mobile app. Walmart has yet to announce website or international availability, but as it's a new feature this may change in the coming weeks. Walmart Sparky also requires users to sign in to their Walmart account, so it can provide personal recommendations based on shopping history. If you're a US customer, you'll find the "Ask Sparky" button at the bottom center of the US Walmart app interface.

This mobile-first strategy contrasts with Amazon's Rufus, which worked across all platforms and markets at launch. It also highlights Walmart's belief in mobile e-commerce, targeting Gen-Z audiences, and allows Walmart to test functionality without disrupting their main website.

6. Current Features vs Future Agentic AI Plans

Right now, Sparky handles product discovery, review synthesis, and recommendations with explanation. It can compare options and help with occasion-based shopping like event planning or seasonal needs.

Based on a statement by Desiree Gosby, SVP of Tech Strategy and Emerging Tech at Walmart, the roadmap includes multi-modal capabilities (text, images, audio, video), automatic reordering based on usage patterns, and service booking for complex purchases. Eventually, Walmart envisions Sparky managing complete shopping journeys with minimal human intervention.

These future capabilities represent true agentic AI—systems that take autonomous actions rather than just providing information, transforming how people interact with retail.

AI in Retail: What This Means for the Future of Ecommerce

The AI in Retail Market is expected to grow from USD $31.12 billion to USD %167.74 billion by 2030, with a CAGR of 32%, indicating Walmart Sparky and Amazon Rufus will not be the only efforts of their kind. AI-mediated shopping is becoming the industry standard, fast, and brands who optimize their SKU's to the unique criteria of each of these models will benefit.

For sellers on Walmart and other platforms, this shift requires new optimization strategies. Software that can help you optimize variants of your listings for multiple platforms, and track their performance in AI conversations. That is where software like Azoma comes in.

At Azoma, we specialize in helping enterprise sellers succeed in AI-powered shopping experiences like Amazon's Rufus, ChatGPT, and now Sparky. Like every other platform, as Sparky continues to evolve, we will adapt our approach to make sure your generated content meets all the benchmarks and appears number one in your category for AI-driven product discovery.

Ready to optimize for AI-driven ecommerce? Contact us to learn how our AI visibility platform can help your products succeed in Sparky and other AI-driven ecommerce experiences.