Сбер и «Сибур» разработают нейросеть для импортозамещения производственного оборудования

Digital assistant will help optimize procurement requests

Sber and SIBUR have started developing a digital assistant based on artificial intelligence (AI). The neural network will allow selecting domestic analogues of production equipment instead of imported ones.

The AI assistant for optimizing equipment procurement is created on the basis of the GigaChat neural network (created by Sberbank). The digital assistant studies the technical documentation of a given item and selects domestic analogues. The main emphasis is on price and quality.

This is a chat for dialogue between a person and AI, working in a co-pilot format, in which a specialist can send requests to clarify the parameters of purchased equipment and materials or select analogues according to the parametric principle. The specialist receives an answer in the chat and, based on it, forms or adjusts an already entered purchase request.
Alisa Melnikova, Director of Digital and Information Technologies at SIBUR

The neural network is planned to be integrated into the companies' business processes. The developers assume that their product will help to fully or partially control the equipment procurement process.

Cloud technologies will not be used to launch the AI assistant — this will preserve information. Servers with software for the neural network will be owned by the company.

The second phase of development is scheduled to be completed in 2025. The estimated cost of the neural network is several tens of millions of rubles.

Earlier, www1.ru reported that the GigaChat neural network from Sberbank will "go" to the university.

Read materials on the topic:

Sber Extends Access to GigaChat Chatbot in Telegram for All Users

The First Real Russian Competitor to ChatGPT? Sber Presents GigaChat

"Spoiler": Sber's Neural Networks Kandinsky, SymFormer, Salyut and GigaChat Wrote a Track and Shot a Video

Sources
CNEWS

Now on home