How to build AI Workflows with long waiting times without draining memory?

Answered

How are we supposed to build AI apps on Xano if requests to large slow models (like OpenAI’s o3-pro or o1-pro) take several minutes to return a response, and after any waiting times over 2 mins, Xano starts draining out lots of memory?

I'm not even talking about multiple llm requests, just 1 that takes probably 5 or 10 minutes to complete, this, multiplied by a large a amount of users doing it at the same time.

Is there a recommended way to handle long-running AI tasks without hitting memory issues?

9 replies