Voice is a technology that's been rapidly expanding, fueled by the growth of smart speakers and assistant apps in smartphones. Despite vast adoption, the market is slow to adapt use-cases that are more complex than setting up an alarm clock. Relying only on audio for input and feedback limits customer trust and makes transactional use cases hard to achieve.

We were presented with a challenge to build a showcase product for Samsung's voice assistant Bixby that would demonstrate a strong use case for voice commerce.


Tapping into Customer Habits

We knew that to build a successful use case for voice in e-commerce, we need to tap into existing customer habits and deliver convenience through voice interactions.

In the last years, we've seen a booming industry of grocery delivery. Major retailers in Europe already offer this service to millions of customers every year.

Online grocery delivery is a multi-step process - apps for grocery ordering often deliver a shopping list feature, where customers can note down their ongoing needs and convert it into a populated shopping cart once it's ready.

Building shopping lists is an ongoing process that fits perfectly well into Samsung's ecosystem, with Bixby being present in various devices, including the latest small fridges. By building a solution where a customer can ask Bixby to add items to their shopping list, we're reducing the amount of effort required as they don't have to use their smartphone.

Well-defined Customer Journey

Customers can gather items on their shopping list throughout the week using only voice interactions with, for example, their smart fridge. Once they want to proceed and order their groceries, they can use their smartphone to review the shopping list and finalize the transaction.

By shifting modalities, we let customers make the final decision on a device that they already trust enough to make purchases. The presence of a screen makes it also easy to review the shopping list and make any changes required.

AI-based Personalization

It's critical to know each customer's habits and consider them in the customer journey to deliver a smooth experience.

As an example, customers will rarely ask their voice assistant for a specific brand of milk but rather for milk in general. They have their preferences that our application needs to learn as well.

To avoid asking them for clarification every time they ask to add milk to their shopping list, we integrated with a personalization system that remembers their choices and provides suggestions based on their purchasing habits. That makes building shopping lists more natural and also allows for faster checkout.


We presented the concept during Samsung's annual Developer Conference with a lot of positive feedback from both people interested in new technologies as well as those working in retail.

Let's Do a Project Together.

Leave your information below

or contact us directly


Upside Lab sp. z o.o.

VATID: PL6762565519

Długa 74/4

31-147 Kraków


Rheinsberger Str. 76/77

10115 Berlin


New Business.