Blog & news 4 min read

Serhii Balashuk presented integration architecture for a large enterprise

On October 8, 2024, Intecracy Group hosted an offline Intecracy Solution Day dedicated to building integration architectures for connecting legacy and modern enterprise systems.

On October 8, 2024, the Intecracy Group consortium hosted the highly anticipated offline event, Intecracy Solution Day, which brought together leading CIOs, system architects, and technical specialists from Ukraine's largest enterprises. The central topic of discussion was the integration architecture of a large enterprise. Participants focused on finding effective solutions to ensure reliable data exchange between legacy systems and modern digital platforms. The key speaker of the event was Serhii Balashuk, a leading integration solutions architect, who presented a detailed analysis of modern technological approaches and practical case studies.

The event was led by IQusion, an Intecracy Group member.

Bridging the Generational Gap in IT Infrastructure

Large enterprises operating in the market for decades typically possess extremely complex and heterogeneous IT infrastructures. On one hand, they rely on time-tested monolithic systems that stably perform critical transactional functions. On the other hand, the business demands rapid deployment of mobile applications, real-time analytics, and flexible cloud services.

Attempting to directly connect these two worlds through traditional point-to-point integrations quickly turns the infrastructure into a chaotic web of connections that is virtually impossible to scale or securely update. Serhii Balashuk broke down in detail why direct integration is a dead-end path and proposed the concept of creating a unified integration layer that acts as a reliable mediator between different generations of technologies.

Technology Stack: API, Kafka, and MDM in Action

To build a stable and scalable integration environment, Intecracy Group architects recommend using a combination of proven tools, each addressing a specific class of tasks. First, this involves standardizing interactions using APIs and implementing an enterprise API Gateway. This component provides centralized access control, authentication, request routing, and protection of internal systems from overloading.

Second, for processing large event streams and asynchronous data exchange, the distributed system Kafka is indispensable. It enables the implementation of an event-driven architecture (EDA), where systems interact by publishing and subscribing to events, guaranteeing data preservation even in the event of temporary failure of one of the consumers. Third, a critical element is the Master Data Management (MDM) system. Without an MDM system that provides a single source of truth for reference data on customers, products, or contractors, any integration will only lead to the rapid spread of errors and duplicates across the network.

“When designing an integration architecture for large-scale systems, we always face a trade-off between data transmission speed and guaranteed consistency. Using an API Gateway allows us to create a clear and secure interface for external consumers, completely isolating them from the complexity of the internal infrastructure. However, when it comes to transferring millions of transactions between legacy monoliths and new microservices, synchronous requests become a bottleneck. This is where Kafka comes to the rescue, acting as a powerful load buffer. It allows for asynchronous event accumulation and processing as consumer systems become ready. And the implementation of MDM guarantees that all these systems speak the same language, avoiding data conflicts at the business logic level,” noted Serhii Balashuk during his presentation.

Practical Trade-offs and Architectural Decisions

During the panel discussion, the offline meeting participants discussed the practical aspects of implementing these technologies in detail. Specifically, significant attention was paid to the Change Data Capture (CDC) pattern, which allows tracking changes in legacy databases without creating additional load on core transactional processes. Changes are read directly from database transaction logs and instantly published to Kafka for further processing by other services.

This approach significantly extends the lifespan of legacy systems while providing modern services with real-time access to up-to-date data. The speaker also emphasized the importance of API versioning and using a Schema Registry in Kafka, which helps avoid failures when updating individual system components.

Strategic View on Infrastructure Evolution

Building a modern integration architecture is not just about replacing old technologies with new ones; it is a strategic step toward creating an enterprise digital platform. Utilizing a modular approach and clearly defined integration layers allows companies to carry out gradual modernization (the so-called Strangler Fig pattern), where legacy functions are replaced step-by-step by microservices without business interruption or risk to critical operations. The Intecracy Group consortium continues to actively share its experience and help Ukrainian enterprises build flexible, resilient, and scalable IT landscapes capable of quickly adapting to changing market demands.