Guest Post: How Microsoft Azure Helped Bismart Build a Magic Mirror to Reinvent Clothes Shopping

Sign in to queue

Description

Generic Image

The following is a guest post by Cillian Shields, Content Manager at Bismart.

The face of retail is changing, thanks to the incredible leaps in technology we've made in recent years. The advent of the internet has revolutionized how people buy clothes, making online shopping the hottest trend. The next step is revolutionizing the in-store experience by adding intelligence.

Enter the Bismart Magic Mirror, which aims to offer consumers a new shopping experience. This will leave customers more satisfied in the store, as they will be shown only perfect, tailor-made choices suited to their tastes. No more shuffling through endless racks of shirts that aren't quite "you": The Magic Mirror gives you exactly what you were looking for, without you even having to go look for it.

How did we create such a powerful game-changer for the world of retail? Intelligence. But if we are to expand on this, we look no further than applying some creativity to Microsoft's Azure Cognitive Services.

Picture yourself walking into your favorite store and the Magic Mirror greets you by your first name as you enter. It knows what you last bought in the store and has an idea on what you might be looking for this time, too. It knows what you like and what you don't like. And it uses all of this to create the best customer experience for you by acting as your own personal shopper.

Application Program Interfaces

How did the Magic Mirror first recognize you as you walked into the store? We were able to build our own Bismart algorithms on top of Microsoft's Face Recognition application program interface (API), allowing this intelligent system to get to know its customers and remember them for the next time they come in. The Magic Mirror gets to know your tastes and preferences through your past purchases, and it aims to use massive data sets to predict what other options you may be interested in.

But also, what if one day you're not quite feeling yourself and need a pick-me-up? The Magic Mirror can recognize emotions in consumers, thanks to the emotion-recognition API, and identify your exact emotional state, broken into eight categories, down to two decimal points.

It recognizes happiness, surprise, neutrality, contempt, anger, sadness, disgust, and fear. In fact, you can test this technology out at any time on Twitter by posting a selfie using the hashtag #bismarter, and you'll receive a breakdown of your emotional state in the photo from Bismart, using our emotion-recognition technology! The Mirror then uses this information to offer you a recommendation based on what you'll probably enjoy best.

Offer a recommendation, you ask? How is the Magic Mirror so knowledgeable? Microsoft Azure also offers a Recommendations API that can add a new dimension to the service you offer your customers. With this technology from Microsoft, we were able to craft an intelligent system interface by combining it with our own algorithms. As a result of these, the Magic Mirror can recognize what items are often bought together, so if one customer is buying a particular shirt, he or she would likely be interested in the pants to go with it.

Recommendations API personalizes the customer's experience by remembering his or her trends, styles, and preferences, and makes recommendations based on this knowledge. This isn't only a tool that makes shopping easier, faster, and more convenient for the customer; it's also a very useful tool to let customers discover more of your products. Once customers see the value in the intelligence of the Magic Mirror, they can be confident that the recommendations they receive from it will be tailor-made just for them, and will very likely commit to purchase.

Lastly, the Content Moderation API helps to run the interface of the Magic Mirror. You can interact with the mirror by speech, text, or image recognition, and the mirror responds in kind. To power this, Microsoft's incredible Content Moderation API lets the Magic Mirror enhance and detect images and video, meaning browsing through a store's catalogue on the mirror couldn't be easier or more enjoyable for the customer.

In this new age of data and information, the extent of possibilities is still just being discovered. Zooming out and taking massive sets of data into account can let you identify patterns, and these patterns can help you solve problems. Big Data can solve problems that up until recently humanity believed simply couldn't be solved.

The Magic Mirror is one example of data being used in a retail environment. At Bismart, we have also found ways to apply our algorithms to government services, healthcare, banking, and myriad other ways. We have developed Traffic Fatalities Prevention services, a Social Budget Planning governmental service to help administrations identify those in need of help to provide it to them, and we can even predict crime.

All our products and services are unique to fit the needs of our clients, be they from the public sector or private. If you're interested in finding out more about what your data can do for you and bring your operations to another level, get in touch with us today.

The Discussion

Add Your 2 Cents