Reduce Manual

Effort

The automated approach reduces the manual effort required to generate, validate and check the product descriptions and ensures that the resultingcopy doesn’t have any grammatical or spelling errors. A trained NLP model generates the descriptions using only the product images as inputs,by detecting the key features of the product and describing them appropriately.

Card image cap
Card image cap

Types of

Copy

The model has been trained to generate different types of copy for the product, including its name, a short description and a longer description that builds on top of the short description by elaborating on the different elements captured in the short description. This ensures that all product related copy on the site is generated automatically and increases the speed with which the client can refresh its online inventory.

Consistent

Style

As the model was trained on the existing samples of the merchant’s product descriptions, it ensured that the automated descriptions were in line with the merchant’s style guide and brand positioning; this reduced the time and effort spent on onboarding new copy writers. Existing manpower was deployed to reviews and to generate copy for products that the model wasn’t yet trained on.

Card image cap

Convolutional Neural Networks

A bespoke machine learning model based on Convolutional Neural Networks was deployed using TensorFlow and trained on 8000+ training images across 200+ cracked and normal devices.

Efficiencies Delivered

The deployed model was able to automatically process more than 80% of the incoming inventory, resulting in a 90% reduction in processing time.

Deployment

The resultant TensorFlow model was deployed at three different warehouses for the customers for processing mobile phones; training is underway to detect cracked tablet screens next.

Get In Touch