Pocket_01Pocket_01

Pocket

Capstone 2017

Pocket is a family of products that enable small businesses to better engage with their customers by leveraging AI to track and aggregate data.

Collaborative design project between Daniel Smitasin & Justin Babikian. 

Pocket_02Pocket_02

We spoke with business owners and observed how their customers discovered and shopped. What we found was an opportunity to homogenize devices used for CRM and create a new interaction model for storefronts.

PocketPOS_01PocketPOS_01

Point of Sale

We started looking at the point of sale as the pivotal centerpiece. Looking at just building the bare essentials we started with a device that could read cards and other cashless payment methods. The screen was built from a need to create an affordance to guide the user to the payment process. After testing angles with a few prototypes we concluded with a 30-degree angle to position a camera for face recognition payment.

DSCF1260-2DSCF1260-2
PocketScale01PocketScale01

As businesses grow so must their tools. We wanted our product to scale with the business and so we created "Plus" a larger point of sale in which the card reader nests in. The Plus provides a feedback monitor, more processing, better communication devices, and receipts. 

PocketScale02PocketScale02
ComputerVisionRenderComputerVisionRender

We wanted to leverage cameras and sensors to track and aggregate data across the store. We started with placing cameras on the point of sale but quickly realized that we had to expand our products to allow for scanning of objects. To solve this we designed two products that would fit into our system. A hand-held wand that would scan inventory and cameras that could recognize objects through machine learning and digital twins. This would allow our system to quickly scan products through the checkout line, provide customer data such as floor heat map, and track customer behavior. These insights will allow business owners to create unique experiences catered towards their customers.

scannerscanner