20 August 2020, Moscow, Russia.M.Video-Eldorado Group, Russia's largest consumer electronics retailer and part of the SAFMAR Group owned by Mikhail Gutseriev, has transferred its hypothesis testing, model training and testing, as well as launching of data science services to Yandex.Cloud. The move enabled the Company to increase more than tenfold the number of simultaneously piloted data analytics projects in just a few months.
M.Video-Eldorado Group has transferred its data-office product development and operation to Yandex.Cloud. The COVID-19 pandemic accelerated the retailer’s transformation into an online business, transition to smart services and automated operations, requiring a rapid scaling of the Company’s data centre capacity. Cloud solutions ensure reliable information storage and access to high performance computing technologies while also enabling transparent cost control. The transition to Yandex.Cloud has helped the Company to triple the ML-development resources and run about 100 pilots and tests at a time.
“In the two years since the opening of the Data Science Centre, we have significantly expanded our competencies, with our development volumes and headcount growing by an order of magnitude. Today, M.Video-Eldorado’s data office is continuously developing and rolling out a number of automation and efficiency enhancement solutions in such areas as customer communications, personalised offers, product range management and promo planning. This growth requires a flexible approach to our computing capacity. With cloud technologies, we can scale up our infrastructure literally within an hour and scale it down almost as quickly when it is not used, saving on major investment in purchasing and supporting our own equipment, said Kirill Ivanov, Head of Data Office at M.Video-Eldorado.
The Company is actively using graphics processing units (GPU) having switched to a computing power consumption model based on specific project needs. To allow for rapid data processing, we have connected our data office to Yandex.Cloud’s data centre through Yandex Cloud Interconnect. Building our own data centre to support M.Video-Eldorado’s data office needs would have considerably increased the project timeline and capital expenditures.
Since 2018, M.Video-Eldorado Group has been using machine learning tools and data analytics in online marketing, e-commerce, logistics, as well as financial and HR management to boost sales and improve the efficiency of business processes. Our projects are mostly aimed at developing recommendation services helping to understand customer needs and extend personalised offers. In addition, they aid in identifying the most popular products, determining the required headcount of retail personnel, and estimating the scope of logistics services and optimal prices. On top of that, M.Video-Eldorado uses AI to develop solutions based on video data analysis to better assess retail performance, reduce the load in certain areas, and offer a more personalised customer experience.