Sberbank adds machine learning-based search in Sberbank Online app

Maria Nikolova

The machine learning algorithms offer personal search results based on the users’ behaviour.

Sberbank Rossii PAO (MCX:SBER) is beefing up the capabilities of its Sberbank Online app via the introduction of machine learning-driven search.

The app search on the main screen allows users to search for contacts for transfers, organisations for payments, app functions and banking products. Now, thanks to self-learning algorithms, search results are adapted to clients’ requests, which allows them to open the needed screen in the app. When users start to input a word in the search, they are shown options that they are very likely to be looking for (for example, address book contacts that the user interacts with most often).

The program analyzes users’ activity both in the app and search itself. The more a client uses the app, the more accurate the results are.

Even before entering the first symbols, users can see recommended operations suggestions – the system predicts what clients will look for on the basis of machine learning. The recommended operations show relevant payments and transfers that were made previously not only in the app, but also at offices, in ATMs and on the website.

The neural network forms recommendations based on popular contacts and operations, taking into account the period of activity and other conditions (for example, as usual the search offers to pay bills at the beginning of the month or transfer money after the client receives their salary). This function is currently available only for iPhone users, and Sberbank says that it will also be available in the Android version in the future.

Thanks to the integration with Sberbank’s natural language processing platform, the search can now understand natural language requests and correct typos, and offer clients relevant answers. Clients can ask questions using words in context or everyday language (for example, ‘How can I apply for a loan’, ‘Where can I find card details’ and so on).

Earlier this year, Sberbank announced the launch of an AI-based chatbot that will help users of electronic trading system Sberbank Markets. The new solution will initially consult clients about FAQ related to the work of Sberbank Markets. In the future, the chatbot is set to gain extra capabilities. It will inform users about trading developments, as well as about perspectives of entering into deals with various asset classes.

Read this next

Inside View

DTCC’s Systemic Risk Barometer Survey found 2024 US Presidential Election as a top risk

U.S. political uncertainty, particularly regarding the 2024 Presidential Election, has emerged as a key risk, with 51% of respondents highlighting it as a major concern. This reflects the potential impact of election outcomes on market conditions and the industry.

Executive Moves

Options Technology promotes Laura McCann to CFO

“Laura’s promotion to CFO is the next stage in our long-term strategy of building a world-class finance team servicing the global business from our Belfast office. Back in 2016, Jon took on the challenge of laying the groundwork for that vision. Laura has been an integral part of the strategy from day one.”

Digital Assets

Thailand’s crypto economy under the spotlight: a report by HashKey Capital

“I’m excited by the rapid expansion of Thailand’s Web3 sector. With over 3 million overall crypto users and 600% growth in the market in recent years, the dynamism in our DeFi and NFT sectors is clearly evident. Thailand is increasingly becoming a hotspot for digital nomads, drawn by our crypto-friendly policies, affordable living costs, vibrant food and beverage culture and diverse cultural landscape.”

Retail FX

Webull Australia offers 5.4% yield on uninvested cash

“US dollar money market funds are heavily regulated, meaning client funds are managed in a safe, reliable and trusted environment, which is of critical importance to us, and continues to remain top-of-mind for our clients.”

Digital Assets

Bybit welcomes Ethena’s USDe, a decentralized stablecoin utilizing delta-hedging staked Ether

“Our collaboration with Ethena Labs represents our commitment to solving some of the biggest challenges in crypto today, not least, the creation of a decentralized stablecoin. The integration of USDe on Bybit expands our stablecoin offerings, providing our users with an array of uncorrelated solutions accessible from our Unified Trading Account.”

Market News

Unravelling the Yen Surge and BoJ Policy Speculations Impacting USD/JPY

The recent downturn in the USD/JPY pair due to the yen’s strength, driven by speculation about the Bank of Japan’s potential tightening of monetary policy.

Digital Assets

Himalaya Exchange customers seek release of frozen funds from DOJ

FormerFeds, a corporate defense and litigation service provider, has filed a lawsuit against the U.S. Department of Justice (DOJ) on behalf of over three and a half thousand Himalaya Exchange customers.

Digital Assets

Nubank, Circle, and Talos join forces for crypto adoption in Brazil

Nubank, the Brazilian neobank backed by Warren Buffett’s Berkshire Hathaway and Softbank Group Corp, announced new partnerships with cryptocurrency firms Circle and Talos.

Metaverse Gaming NFT

Flare onboards Ankr, Figment, Restake, and NorthStake as validators

Flare, an EVM smart contract platform known for its focus on blockchain data utility, has announced a major step in its development. The platform has onboarded leading infrastructure providers, including Ankr, Figment, Restake, and NorthStake.

<