Our client is a New Delhi-based startup that wants to revolutionize the retail apparel industry by setting up kiosk systems in their physical showrooms. These android based kiosks intend to help the customers choose and order products that are currently not available in the store. The kiosk also takes in the information of the customer and delivers the products to their desired address once they are restocked.
The client needed the kiosk to work on a single application mode which means users can not use this kiosk for any other app than the client’s application.
The search feature in the app must be equipped with an AI feature where the user can upload any images of shirts, t-shirts or trousers for the system to recognize the colour, pattern, and style of that apparel. It should also be able to showcase options that match with the search made by the user.
We customized Android OS and implemented a lockdown mechanism to achieve a single application mode. To ensure flawless customization, we worked closely with the client’s kiosk vendor/manufacturer who ultimately installed the customized OS on the devices.
Since this is a SaaS-based product, the major question was to customize the app’s UI to match every customer’s style. We crafted and integrated a server-driven UI for the whole application to ensure that each user gets a personalized and seamless experience.
Technically, the tagging engine uses a custom deep learning library that is also trainable for defining tags, specific to use cases while the user tries image-based search. The system can assign relevant tags to products making it an extremely smooth and personalized experience.
