Integration Flow

Integration Flow is a visual orchestration engine designed for long-running and large-scale processes. It allows you to connect rules, APIs, databases, and data transformations into one executable backend workflow that runs asynchronously in the background.

It was built for workloads that exceed the practical limits of real-time execution.

Decision Flow is optimized for immediate, real-time decisions — typically one request in and one response out.

Integration Flow is optimized for processes that take longer to complete. Instead of prioritizing instant responses, it focuses on reliable execution of larger workflows running in the background.

The two are complementary rather than competing.

Integration Flow was designed specifically for long-running workloads. Unlike real-time execution flows that risk timing out after a short execution window, Integration Flow runs in the background and can continue processing for up to 24 hours if needed.

This makes it suitable for:

Traditional real-time orchestration expects an immediate response. That works well for frontend decisions and lightweight processing, but complex operations can quickly run into timeout limitations.

Integration Flow uses background execution instead. The process continues independently of the original request, allowing complex workflows to finish properly even when they involve large datasets or slow external systems.

Yes. Integration Flow is specifically suited for database-heavy workloads such as retrieving large result sets, processing records sequentially, synchronizing systems, or updating external data sources.

This makes it ideal for backend operational processes that would be difficult inside real-time execution flows.

Yes. Integration Flow supports REST API integrations directly inside the Flow. Because execution runs in the background, it is much better suited for slow or complex external integrations than real-time orchestration.

Yes — this is one of its primary use cases. Integration Flow was designed specifically for high-volume and large-scale processing where many records, requests, or operations need to be executed as part of one workflow.

Move to Integration Flow when your process stops behaving like a real-time decision and starts behaving like backend processing infrastructure.

Typical signals include:

Use Decision Flow for immediate decisions.

Use Integration Flow for long-running orchestration.

Keep orchestration readable and modular. Separate retrieval, transformation, evaluation, and persistence into clearly defined stages instead of mixing everything together in one massive workflow.

Well-structured Integration Flows are easier to debug, safer to modify, and much easier for teams to understand over time.

Usually no. As a best practice, keep business decision logic inside reusable rules like Decision Tables, Decision Trees, or Decision Flows.

Let Integration Flow focus on orchestration:

This keeps your architecture cleaner and your logic reusable across both real-time and long-running workloads.