Director of new projects of Ensi Sergey Melikhov connects the high level of AI penetration in e-commerce with a combination of several factors: "In e-commerce, most processes are digitalized or even completely take place in information systems. This means that improving them with the help of artificial intelligence technologies is easier and more natural. In addition, e-commerce companies in most cases initially developed as IT organizations and managed to accumulate philippines whatsapp number database enough expertise. It is worth noting here that the hype around AI, machine learning (ML) and large language models (LLM) technologies happened after the release of popular products like ChatGPT and Midjourney, but these approaches have been used in IT for a long time. Finally, e-commerce companies accumulate big data, and, having this asset, have long been engaged in extracting added value from them. And in traditional retail, working with data is almost always a project for implementing innovations. As a result, for the e-commerce sphere, AI and ML are just technical tools that can be used in any zone. These technologies have improved and become cheaper, and the world is now experimenting, choosing economically demanded niches for them. It is quite possible that after some time we will stop discussing them, because they will become ordinary tools, like some object-oriented programming."
"The e-commerce sphere has long been implementing artificial intelligence, but the last few years have seen a significant evolution in this area. Instead of creating separate models for each task, companies began to actively develop universal multimodal engines that can effectively work with various types of data, such as text, images and video. However, such technologies require significant investments in equipment and specialists with the necessary skills to work with large models and volumetric data. Therefore, access to advanced AI technologies remains the prerogative of large companies, some of which are in the e-commerce sphere. It seems obvious to me that the implementation of neural network technologies will continue to grow. A lot of new opportunities will appear in the field of content generation. This will also change promotion methods - it will be possible to create super-personalized marketing campaigns. For buyers, I expect further development of personalization and approaching the one-click shopping format. The better the content is made and the recommendation systems are set up, the less actions the buyer will need to take to find the right product. And if we really fantasize, then perhaps sites like Avito will be able to immediately recommend a product based on a need - say, "a gift for a woman, at such and such an age, for such and such an occasion." And at the same time, they will take on the function of market research, choosing the best product among similar ones. And this, by the way, means that the current advertising model may lose its relevance."
Avito's Director of Data Andrey Rybintsev called the use of AI in e-commerce a long-standing practice:
-
- Posts: 422
- Joined: Thu Jan 02, 2025 7:50 am