To build solutions for cash replenishment of Sampath Bank PLC Automatic Teller Machines’ (ATM) and to recognize withdrawal patterns.

Challenge: Sampath Bank had a lot of operational complexities in their ATM environment with 400 plus machines around the country. Therefore, KBSL had be mindful of major challenges which came its way- such as stakeholder coordination, transportation of money and maintenance of ATM.


The solution has been designed by employing one of the branches of Artificial Intelligence (AI) known as Artificial Neural Network (ANN). ANN model has been exclusively selected for the project owing to the following reasons.

  • Capability of self-learning
  • AI is being the future of Business Intelligence
  • Openness of the solution
  • Capability of future development / amalgamation or integration of any upcoming platform
  • Robustness which will enhance the total banking experience at an intelligent level, a stepping stone of new banking era

Furthermore open source being the principal platform deviser, it provides all the parties (Users and Developers (KBSL Data Labs) and independent customization ability for better user experience.

We then coined the ANN solution which consists of,

  • Prediction for Cash Demand
  • Currency Notes Denomination
  • Cash Replenishment Frequency Identification
  • Business Intelligence and Further Insights
  • Visualization of complex binary information to easily understandable charts and graphs  

There are three sub sections of the overall process which simplifies the architecture and increase the integrity of different steps.

  • Data Extraction and Transform
  • Machine learning Algorithm Execution for Prediction
  • Business Intelligence and Further Insights