Hungary plans to detect anomalies in e-government services with Artificial Intelligence

The Responsible Organisation 

IdomSoft is a Hungarian state-owned company and a major actor in Hungary’s e-government development. They are responsible for the deployment of a wide range of different digital services, from drivers’ licenses to identity cards and voting systems. 

The problem 

E-government services are essential for citizens as they allow them to access essential information and benefit from different administrative services in a simple and time-efficient way. However, digital services are often vulnerable to disruptions and inconsistencies, causing security and operational risks, besides leading to frustration and disruptions for users. Additionally, system administrators face the continuous challenge of monitoring massive amounts of data generated by digital services and identifying potential issues in advance. 

The solution and its implementation 

The solution leverages advanced Machine Learning technologies to ensure the optimal performance of critical e-government services for citizens in the country, improving their reliability and efficiency. The system uses models for transforming non-structured into structured data, employs deep learning on neural networks to predict events, and implements Machine Learning Operations (MLOps) methods to manage the data pipeline and the model. The solution is also based on the concept of "Machine Learning as a Service" (MLaaS), providing a user-friendly interface for custom data analysis. 

This solution is expected to: 

1. Ensure the smooth operation of services minimising the risk of service corruption,  

2. Predict future load by anticipating potential bottlenecks and resource constraints,  

3. Identify misuse through the detection of anomalies and suspicious activity.  

Expected benefits 

The solution is expected to demonstrate several benefits for both the operational and quality aspects of e-government services deployment: 

• The solution reduces operational costs and ensures a more efficient use of resources, by lowering the costs of maintaining IT systems.  

• The simplification of the maintenance and troubleshooting of the services makes the digital services run more smoothly. 

• The automation of system administration and monitoring tasks also diminishes the workload for the services’ administrators.  

• At the same time, the solution improves the services’ reliability as it ensures steady and consistent performance of e-government services, which ultimately will contribute to better service delivery and improved user experience for citizens. 

Current challenges 

Along with the described benefits, some challenges have been identified until today: 

• From an organisational perspective, the reduced need for administrative workforce of the system is presenting the need to adapt organisational structures.  

• Additionally, issues related to the management of personal data arise, leading to the need of anonymising data to ensure compliance with the General Data Protection Regulation (GDPR). 

• Idomsoft Zrt. website