RuProject for MTS Bank – Implementation of an information system for forecasting the amounts and timing of replenishment and cash-in-transit for self-service devices
The project is designed for the implementation, support, and functional development of an information system that forecasts the amounts and timing of replenishment and cash-in-transit for self-service devices using neural networks
Key Functional Modules:
- Use of neural networks (autoencoders, word2vec, GloVe, etc.)
- Selection of the optimal forecasting method for a specific device
- Continuous retraining of models to improve accuracy
- Breakdown by cash operating units
- Editing and approval of proposed amounts
- Accounting for all types of self-service devices, insurance limits, and nearby devices
- Display of balances, withdrawals, and forecasts
- Highlighting of deviations and risks
- Ability to respond promptly
- Differentiated calculation system
- Accounting for average cost and fee directories
- Creating and configuring custom roles
- Assigning functions and access rights
- Exporting/importing roles
- Management of calculation parameters (funding rate, rounding, operating modes)
- Accounting for public holidays (support for multiple calendars)
- ProView, OpenWay, processing, and others
- Receiving cards, balances, and turnovers
- Creating plans based on real data
Additional Features:
Calculation of the optimal loading amount and banknote composition
Accounting for insurance amounts and funding
Automatic sending of requests by email using a template
Cost optimization: cash, insurance, labor
Automatic inclusion of devices in the plan (based on criteria)
Advantages of the system:
Результат
- Improved accuracy of cash flow planning and security
- Optimization of cash collection and cost reduction
- Flexible customization to suit the bank’s business logic
- Relevance ensured through continuous development and support
Want to discuss a joint project?
Be our partner and together we will achieve your expectations