Our Hyper-Tagging Engine technology is at Velou's core.

Weaving together AI, machine learning, natural language, processing, and computer vision, we’re able to significantly improve the product data and make it easier for shoppers to find what they want.

Image Analysis

First, we scan each image with Computer Vision to single out the product and then analyze it to identify hundreds of variations of product features. The measurement from the waist to hemline tells us the length of a dress. The shape and curve of a neckline differentiates a t-shirt. We can increase the search attributes for an entire product catalog tenfold.

Dress Style: fit and flare, gown
Embellishment: keyhole back, pleats
Neckline: plunge v-neckline
Closure: single button, hidden zipper
Sleeve style: illusion, bishop
Fit: fitted top, loose skirt
Sleeve Length: long
Material: silk charmeuse, mesh
Color: Red
Hemline: hi-lo
Pattern: solid
Dress Length: floor length

Color Analysis

Our AI-powered color analyzer identifies the various colors in the product, while ignoring any colors in the rest of the image. Then it prioritizes color tags based on the percentages present in the product. So when shoppers include a color in their search, they get accurate and meaningful results.

Text Analysis

We use Natural Language Processing to analyze the text in product descriptions and the existing metadata to develop a deeper understanding of the products and generate a thorough list of tags.

Style: crossbody
Color: peach
Closure: magnetic flap closure
Embellishment: filigree corner protector trim
Handle Style: top handle, chain strap
Material: grained leather, goldtone metal
Occasion: semi-formal, luncheon
Shape: square

AI + Human Intelligence

Our AI models continually monitor for search accuracy. Together with our industry experts, they make sure each product is optimized for discovery.

Let’s connect

Keep up with Velou news and product releases.