How neural networks
can help a business

Create ads

Neural networks help a person create advertising for a business. Yandex uses the YaLM language model to generate ad texts and titles, and the neural network that creates images in Masterpiece now illustrates ads from Yandex Business users.

Fight against fraud

Neural networks are able to recognize fraudulent transfers. For example, Alfa-Bank has implemented an algorithm that monitors transactions. The neural network recognizes suspicious transfers (but does not block safe ones).

Predict fashion trends

For example, the Heuritech service helps clothing brands predict what will be in fashion in the next 12 months. It analyzes photos shared by users on social networks, and then algorithms group images by style, region and a number of other parameters to understand what exactly people like in different countries. This allows you to create models that are more likely to be in demand.

Simplify user registration

Car sharing and other services use neural networks to speed up the process of identifying new users: algorithms recognize and verify their documents. And neural networks also help to fight against scammers who use other people's accounts. So, before giving the driver access to the car, the service will ask you to take a selfie, and the algorithm will check the photo against the database.

Testing user interfaces

The Autify platform helps engineers test online stores and other user services. The neural network creates test scripts that simulate consumer behavior. This helps to identify weaknesses in the interface and predict possible failures to help developers fix them.

Russian version