Back to Blogs

Data Scale: Building and Launching the Corporate Integration Bus

Integration Without Chaos: How We Built a Platform for the Digital Bank of Tomorrow

The IT landscape of a modern bank is a maze of dozens of disparate systems that all need to speak the same language. Core banking systems, remote banking services, CRMs, mobile apps, scoring platforms, regulatory bodies—each new system that connects through a separate point-to-point „bridge“ to every existing one turns the architecture into a tangled web of wires. It’s expensive, inflexible, and risky.

Data Scale offers a different approach: a single, managed integration layer.

We built the Corporate Integration Bus — an infrastructure platform of the Enterprise Service Bus class that acts as the bank’s central nervous system. Instead of hundreds of point-to-point integrations, we provide a single point of entry, unified routing rules, and full transparency and control at every level.

„An integration bus is not just a technical solution; it’s a strategic platform that enables a bank to respond flexibly and swiftly to market changes,“ notes Roman Soloviev, CEO of Data Scale.

Why We Decided to Build Our Own Product

The market for integration solutions is far from empty. However, we identified a core problem: off-the-shelf ESB products are either too heavy, demand years of customization, or fail to account for the specific realities of the Russian banking sector.

We took a different route. Our product is built on:

  • Real operational experience — we understand how banking systems work from the inside out.
  • A modern technology stack — microservices, Kubernetes, and stream processing.
  • A focus on manageability— designed not for „set and forget,“ but for industrial-scale operation with full observability.

This is how we built a platform that doesn’t merely connect systems but becomes the foundation for a bank’s digital transformation.

What Problems the Platform Solves

Our bus addresses the core pain points that every major bank encounters.

  • Centralized exchange — all systems interact through a single interface. No complexity when onboarding new services.
  • Mission-critical reliability — the message broker guarantees delivery, eliminates data loss, supports retry processing, and provides duplicate protection.
  • Freedom to replace systems — want to change your core banking system or CRM? No need to rewrite all integrations. The bus acts as an abstraction—swap modules without touching consumers.
  • Full control — metrics, logs, health checks—we track every message, every route. Diagnosing issues takes minutes, not days.
  • Ready for growth — new digital products are connected via adapters, not by restructuring the architecture. The bank gains speed, not technical debt.

How the Product Works

We built the platform as a set of modules, each solving its own task—yet together they form a cohesive ecosystem.

  • Message Broker (Apache Artemis) — ensures reliable storage, routing, and queue management—the backbone of asynchronous exchange.
  • Adapters — „translators“ for each external system (core banking, CRM, external APIs). They accept requests in the bus format, transform them into the target system’s format, and return results.
  • CDC Service (Debezium) — stream processing of database changes, reacting to events in real time and publishing them to the broker.
  • Observability Subsystem — a unified hub for collecting metrics, logs, and service health monitoring via health-check endpoints.
  • CI/CD and Configuration Management — automated build, testing, and deployment across environments. Management of Kubernetes resources via Helm and Kustomize.
  • Automated Testing Toolset — end-to-end and sanity checks based on Apache JMeter.


Technology Stack

Our technology choices were driven not by trends, but by the demands of performance, fault tolerance, and compatibility with banking infrastructure.

At the core is a Kubernetes cluster, enabling independent scaling of components based on load. The Apache Artemis message broker provides persistent storage and reliable delivery. Stream-based change processing is built on Debezium, while configuration and deployment are automated via Helm and Kustomize within GitLab CI.

The product runs on x86_64 and ARM64 architectures and is not tied to any specific hardware vendor.

Product Advantages

  • Centralization of integrations — the bus creates a unified interaction layer between systems, reducing IT landscape complexity and simplifying integration maintenance.
  • Reduced dependency on specific systems — acts as an abstraction over existing banking systems. Individual modules or external systems can be replaced as needed without changing interfaces for consumers.
  • Reliable data exchange — message broker, queues, retry delivery, deduplication, and idempotent processing minimize the risk of message loss or duplication.
  • Scalability — microservice architecture and Kubernetes enable independent scaling of components based on load.
  • Manageability and transparency — the observability subsystem provides control over service health, visible metrics, logs, and rapid issue diagnosis.
  • Rapid onboarding of new systems — new banking or external services can be connected via adapters using unified exchange and routing rules.
  • Ready for industrial operation — includes CI/CD, configuration management, automated testing, health checks, and operational mechanisms required for enterprise banking environments.

What This Solution Delivers to a Bank

The Corporate Integration Bus enables a bank to transition from fragmented point-to-point integrations to a unified, managed integration platform. This reduces technological risks, accelerates the onboarding of new services, improves the reliability of data exchange, and establishes a foundation for the further development of the bank’s digital offerings.

About the Product

This product is built on hands-on experience with banking IT landscapes. We designed it with the requirements of the financial sector in mind: fault tolerance, manageability, security, and compatibility with existing infrastructure.

Our architectural decisions and technology stack were not chosen in theory, but derived from real-world operational scenarios. Every module—broker, router, adapters, CDC, observability—addresses a specific challenge we have encountered in actual projects.

The Data Scale Corporate Integration Bus is ready for deployment within a bank of any complexity. It does not require restructuring existing systems and allows new services to be connected without modifying already-functioning integrations.

Back to Blogs

Die Website verwendet Cookies für Analysen, Werbung und zur Verbesserung der Benutzererfahrung. Indem Sie weiterhin auf der Website surfen, stimmen Sie unserer Verwendung von Cookies zu, wie in unserer Cookie-Richtlinie beschrieben.