Feature Request: Efficient Handling of Long-Running External Calls in Workflows

Problem:

Some workflows rely on complex prompts sent to models like openai o3, where a single completion can take several minutes. These calls must run synchronously because the response is needed for immediate use in the next steps of the workflow.

Currently, there’s no way to offload or suspend execution during that long wait. Even if post-process is used to avoid frontend timeout, the backend continues consuming compute resources the entire time—just waiting.

This results in significant resource waste and limits Xano’s ability to handle AI-integrated workflows at scale.

Proposed Capability:

Enable native support for deferred synchronous execution, such as:

A way to "yield" long-running tasks while maintaining workflow state

Background or scheduled task queuing with result handoff

Hybrid sync-async patterns where Xano handles the orchestration internally

Why it matters:

Prevents backend compute waste during idle waits for AI completions

Reduces cost for workloads that can’t be split into webhooks or async branches

Supports modern AI integrations (like o1, o3) without infrastructure hacks

Makes Xano more scalable for advanced, production-grade workflows

1
2 replies