Wpis dostępny jest także w języku:
Allegro is working on a new feature for its mobile app. The visual search engine analyzes the item seen in the photo provided by the user and then discovers comparable goods on the allegro.pl website using a machine learning algorithm. The consumer does not have to visit the store and photograph the object. Pictures of goods, such as those seen in magazines or screenshots taken on a phone, are recognized by the visual search engine.
The Allegro app now has a new function that allows users to shoot and share photos of items such as clothing, a bottle of water, or a book. The Allegro tool, which is powered by machine learning algorithms, will then scan the site for products and present the best deals. The function now covers approximately 100 million offers in the most popular categories.
The user is not required to physically go to the store and photograph the merchandise. Pictures of goods, such as those seen in magazines or screenshots taken on a phone, are recognized by the visual search engine. On average, 21 million people visit the Allegro platform each month.
A totally new search architecture and machine learning techniques were required for the visual search engine. A complicated structure lies underneath the user-friendliness. Now, the platform is entering a new age, with the development of a tool based on machine learning algorithms that will make browsing and ordering items easier, according to Wojciech Bogdan, Allegro’s Chief Data Officer.
According to PMR, the e-commerce market in Poland was worth PLN 88.7bn (€19.4bn) in 2020
, growing by 45.1% compared with 2019, according to our estimates. For the period until 2026, our baseline scenario is that the market will continue to expand at a double-digit pace through 2023, and that it will grow by 7-9% per annum in the three years after that, expanding by a total of 89% over this period.