We realised that small businesses are missing out on lots of key data, data that would be able to be interpreted to let merchants know which product is more popular with which crowd. We looked at various methods to enable these small businesses to collect data, but we realised that the most convenient method would be to analyse customer's faces in order to fetch real time data.
What it does
SaleVision is a Point-Of-Sale system that scans customer's faces as they are paying at the counter in order to generate insights and record customer demographics. This data would allow merchants to find out more about their customers in a seamless and unobtrusive way.
How we built it
We used ReactJS for a more flexible front-end. This front-end automatically takes a snapshot of the customer's face every 4 seconds in order to run it against a facial recognition API. The API would return us some customer demographics that we send to an SQL server that formats the data. We would then run some data analysis in order to get insights from the data, such as R-square, X-bar, Y-bar and linear regression.
Challenges we ran into
ReactJS was a beast to deal with and lock down. It required a shift of thinking about how to model our views, but in the end, we are able to use the components we made to easily format and display our views.
Accomplishments that we're proud of
We were able to come up with a working prototype of an app that accurately captures customers details and generates insights from there.
What we learned
We learned that even though we got the initial components of each process up fairly quickly, integrating the systems together took up a huge chunk of time.
What's next for SaleVision
We're aiming to better refine the facial recognition API in order to deliver more data such as ethnicity. The technology is only going to get better with time, and thus, SaleVision will only get better with time. We're excited to see the boundaries we can push with this new Point-Of-Sale system.