We connect SAP, Oracle, AFAS, custom ERPs, payment gateways, and legacy stacks with production-grade middleware — battle-tested on systems that move millions of euros per day, with full audit trails, idempotent connectors, and rollback paths built in. From SAP-AFAS daily reconciliation to high-throughput payment gateways, we build the integration layer enterprise architectures actually need.
In tech, true autonomy strengthens bonds. By transferring knowledge and best practices, we enable independence while forging long-term partnerships, earning the trust of Europe's premier organizations.
Every enterprise has the same problem: finance is on AFAS, the parent group runs SAP, billing lives on Oracle, payments flow through a gateway, and analytics needs everything in one place. Most integration projects start ambitious and end in flat files exchanged by FTP because the underlying middleware couldn't handle schema drift, retry logic, or audit requirements. DNA Solutions has built integration layers that survive — including BridgeSync, the SAP-AFAS middleware automating daily reconciliation for Belgium's national toll operator, and PaySync, the payment gateway running for DriveHub. We design for the hard parts: idempotency, failure recovery, audit logging, and zero-downtime cutover.
From point-to-point ETL middleware to enterprise-wide integration platforms, we cover the full integration value chain — always grounded in real production constraints.
Production middleware connecting SAP, Oracle, AFAS, Dynamics, and custom ERPs. We design connectors that handle schema drift, partial failures, and reconciliation queues — not just happy-path data transfer.
Our BridgeSync platform automates the full daily SAP↔AFAS exchange for Satellic (Belgium's national toll operator), with built-in validation, structured logging, and rollback paths. The same architecture pattern applies to any heterogeneous financial estate.
High-throughput payment and billing systems with audit-grade reliability. We've built infrastructure processing hundreds of millions of euros per month, validated by Deloitte audits, running on event-driven AWS microservices.
PaySync (DriveHub) and the Toll4Europe billing platform are examples of mission-critical systems that combine high volume, financial accuracy, and regulatory compliance — patterns we apply to any enterprise that can't afford a billing outage.
API layers and data connectors that bring modern applications (cloud, AI, mobile) to systems that have been running unchanged for 15+ years. Mainframes, AS/400s, Oracle stacks with lost institutional knowledge — we've integrated them all.
Our engineers spent two decades inside the kind of systems most integrators only read about. The integration layer extracts, transforms, and feeds legacy data to modern consumers without modifying source systems.
Streaming pipelines, message queues, and change-data-capture (CDC) for use cases that need data flowing in seconds, not hours. We design for ordering guarantees, exactly-once semantics, and back-pressure handling.
We've built CDC pipelines feeding analytics platforms (T-Systems Data Lake), event buses for fleet dispatch (Concreteasy), and real-time integration between operational systems where latency translates directly into revenue loss.
Integration projects fail more often than any other category of enterprise software. Here's how we think about the most common concerns.
Because the middleware was built for the happy path. The first time SAP changes a field, AFAS schema drifts, or a network hiccup causes a duplicate transfer, the integration becomes unreliable — and finance teams quietly fall back to spreadsheets. We design for the failure modes from day one: idempotent connectors, structured logging, reconciliation queues for any record that fails validation. Daily reconciliation should be invisible, not a 4-day month-end ritual.
Yes, and that's almost always the right approach. We sit our integration layer on top of your existing systems — extracting via supported interfaces (OData, REST, database CDC, or even structured file exports where nothing better exists) and never touching the source code. The legacy system keeps running unchanged.
Every integration includes validation, reconciliation queues, and audit logs. Failed records go to a queue with full context so they can be inspected and replayed rather than silently lost. For financial integrations, we add daily reconciliation checks comparing totals across systems. We've passed Deloitte audits on billing infrastructure processing €250–300M monthly using these patterns.
Depends on the use case. Daily financial reconciliation = batch with structured logging. Payment authorization, fleet dispatch, fraud detection = streaming with sub-second latency. We design each flow for its actual business clock, not a default architectural choice. Most enterprise estates need both, integrated coherently.
Point-to-point ERP integration: 6–10 weeks for a first flow, then 2–3 weeks per additional flow once the middleware foundation is in place. Full integration platform with multiple flows, monitoring, and parallel-run validation: 4–6 months. We deliver one flow at a time with two-week parallel-run validation before each cutover — no big-bang risk.