About Sberbank AI
Sberbank AI translates data related to client’s financial behavior into a better understanding of how to improve the bank’s services and solutions. The company uses AI and machine learning at every stage of its business processes, from new product development to selecting loan terms.
1 / 6
Automated Decision Support & Scoring
Automated Machine Learning: automated machine learning technologies based on user data.
- The analysis of service quality, call center performance, advertising and marketing strategies.
- Pricing and product development recommendations.
- Automated non-payer identification, overdue loan data collection, individual message generation, and claim statement preparation.
2 / 6
Computer vision: complex user identification system.
- Clients identification by face, behavior recognition, and text data.
- Fraud detection and prevention.
3 / 6
Natural Language Processing
Natural Language Processing: technologies for processing and understanding the natural language.
- Automated voice menu.
- Chatbot for product advice and client’s account balance.
- Automated generation of legal documents.
- Automated processing of complaints and requests.
- News feed analysis to forecast service and product demand.
4 / 6
Automated Machine Learning
Automated Machine Learning: automated machine learning technologies based on user data
- The analysis of clients’ behavior patterns, financial forecasts, and generation of individual offers.
- Clients segmentation by various parameter, including financial behavior and expense patterns, and generating product recommendations.
- Monitoring and adjustment of machine learning algorithms in real time.
5 / 6
Recommender system: technology for the assessment of clients’ interest in company’s products
- Product selection, from loans to paycheck projects.
- Client relations customization.
- Investment recommendations, such as securities selection, toxic asset sale, acquiring, trading, etc.
6 / 6
Speech Analytics: technology for speech recognition and synthesis, such as user voice identification
- Cognitive operator assistant to disclose additional information for client consultations.
- Robotic system to process simple client requests.
- Client identification by their biometric profile (age, gender, and voice recognition).
- Forecasting the probability of purchase, debt collection, etc.