I led the design of Sam's Club's most impactful internal application, a voice-first AI assistant that reduced task time from 38 minutes to 33 seconds saving the organization an estimated $452.4M annually.
Sam's Club associates were losing 38 minutes per customer inquiry due to outdated DOS-based systems, costing millions in productivity and customer satisfaction. Store managers were spending 2 hours daily on manual scheduling processes.
With 600+ stores and thousands of daily customer interactions, these inefficiencies were costing Sam's Club approximately $452.4M annually in lost productivity alone.
I conducted observational studies, time studies, and in-depth interviews with 15 associates and 8 managers across 3 store locations to understand the full scope of operational challenges and calculate baseline metrics for completion of tasks.
During their shifts, Sam’s Club employees were often interrupted by customer questions while completing tasks across departments. Since the only way to look up product info was on slow, outdated back-office computers, it took an average of 38 minutes to provide an accurate answer for a customer.
When customers asked about stock at nearby stores, associates had no seamless way to check. The only option was manually calling other locations, waiting on hold, hoping someone could confirm availability.
Managers spent 2 hours every morning filling schedules by calling an outdated phone system to learn about call-outs. No dashboard, no alerts, no coverage suggestions—pure fire-drill mode.
The team outlined challenge found in research and met those with ways to solve the problem to not only help users needs but business needs. We found many issues but these were the ones that the most sense and what I presented in finding tangible ways to solve massive challenges to increase ROI, improve employee's productivity, and increase customer's satisfaction in the store to improve sales.
Designed a mobile voice assistant that associates could use while staying with customers. Natural language processing allowed questions like "Do we have any Dyson vacuums in stock?" with instant audio and visual responses. Result: Reduced inquiry time from 38 minutes to 30 seconds.
With Ask Sam, the associate can quickly see the quantity and details of the product needed in nearby stores. The nearby feature improved customer service and prevented shoppers from purchasing the item at another retailer.
With Ask Sam, associates are able to quickly view and contact qualified employees to fill in. This is a huge relief for both associates and managers.
Beyond user research, I had to navigate significant technical and organizational constraints that shaped our design decisions.
Using research insights, I developed a voice-first AI strategy that would transform how associates access information while staying connected to customers.
I explored multiple interaction patterns, testing voice-first vs. visual-first approaches through rapid prototyping.
I iterated through multiple wireframe versions, balancing voice interaction with visual feedback requirements. Key decisions included:
Working closely with the evolving Sam’s Club design system, I crafted interfaces optimized for quick scanning and voice interaction. Although it’s rare for external partners to influence internal governance, our team was invited to contribute directly to the new design system. Our work on Ask Sam had gained visibility across teams—and because of its impact, we were asked to help contribute to visual styling guidelines alongside Sam’s Club’s internal governance group.
I conducted usability testing with actual store associates, measuring task completion time and satisfaction.
Ask Sam became the first voice-activated assistant deployed at enterprise scale in retail, transforming how 600+ stores operate daily.
Ask Sam became the template for future Walmart internal tools, establishing voice-first design patterns used across 12 subsequent applications. The success led to my team being retained for additional projects worth $14.8M.
I'd be happy to share research protocols, detailed wireframe iterations, or discuss my approach to voice-first design challenges.
Let's Connect