Walmart Invests in Diverse AI Models with Innovative Wallaby LLM Launch

Walmart is embracing artificial intelligence with the development of its own large language model (LLM), Wallaby, designed specifically for retail applications.

Wallaby is trained on decades of Walmart data, capturing the conversational styles of both employees and customers. Its responses are tailored to resonate with Walmart's commitment to exceptional customer service.

In a recent interview, Desirée Gosby, Vice President of Emerging Technology at Walmart Global Tech, emphasized the importance of thorough internal testing before a wider rollout. “Wallaby is currently undergoing extensive internal testing, particularly among our associates,” she stated. “In the upcoming year, we plan to leverage it further.”

While initial use cases for Wallaby will likely focus on customer interactions, it will also form part of a broader suite of models as Walmart develops new applications.

Given Walmart's scale as one of the largest retailers globally, it is logical for the company to invest in proprietary LLMs. Internal model development can be costly, but Walmart has access to a wealth of customer, employee, logistics, and retail data that enhances the effectiveness of its AI models.

Walmart employs a multi-layered strategy for AI integration, using a platform called Element to incorporate various third-party and proprietary models tailored to specific applications. “This approach helps us manage different models effectively, with Wallaby serving as one of our retail-specific LLMs,” Gosby explained. “Ultimately, our strategy will depend on the problems we aim to solve, which may involve a combination of models.”

During a media event in July, Gosby highlighted Walmart's increasing adoption of technology, mentioning that the retailer has been utilizing GPT-4 for various AI applications, especially with the growing digital engagement from customers on its mobile and web platforms.

Walmart isn’t the only retailer leveraging AI. Amazon has introduced Rufus, a chatbot powered by AWS models, which provides product information and customer reviews.

At Walmart's headquarters in Bentonville, Arkansas, AI has long been a part of employee operations. The company implemented a chatbot for associates to inquire about policies and employee handbooks and integrated AI to enhance supply chain management.

Additionally, AI technology plays a crucial role in Walmart and Sam's Club stores, assisting floor associates with inventory assessments and checkout procedures.

Walmart has also upgraded its generative AI tools on digital platforms. Gosby noted that while the retailer's customer support features have existed for years, enhancements were recently made to its Customer Support Assistant. This chatbot now recognizes customers from the outset and understands their intent.

In a demonstration, customers could naturally express returns, such as "I accidentally bought some toys." The assistant intelligently focused on "toys," retrieving relevant order details while ignoring unrelated purchases, even proactively asking if the customer wants to return or keep the item.

Looking ahead, Walmart plans to introduce more personalized recommendations and tailored homepages for customers. Gosby affirmed that the company will continue exploring how AI can enhance experiences for both customers and employees, with Wallaby expected to play a significant role in future applications.

Most people like

Find AI tools in YBX