Content is an effective solution for driving sales and improving customer experience. AiBUY connects consumers to products on any device, anywhere, uninterrupted.
More likely to make a purchase after watching a product video.
More clicks than a non-video sidebar calls to action.
More likely to buy a product viewing a video in the pre-purchase stages.
Using AiBUY’s proprietary machine learning and computer vision algorithms, we can detect any item through an image or video, making the eCommerce experience as fast and native as possible.
The solution employs microservices architecture and AWS for storing images, vectors, and AI models. Then the pre-trained convolutional network finds and suggests the most likely product category. The convolutional network transforms an image into multi-dimensional vector.
The algorithm compares the image vector with other vectors from the existing product catalog.
Then, the solution sorts vectors based on their similarity values. The vector with the minimal distance is the most appropriate item for the product catalog. So we can assume that this item has the most similarity to the source item. The other, more distant vectors will be also shown to the user as similar ones.