I am building an app using lovable.dev, and it involves integrating with an AI agent workflow on n8n over a production webhook. I am using n8n online.
The integration is working fine and I am able to receive responses from the agent and render them on the app, although sometimes for some complex queries the agent stops due to max iterations, which I have not been able to stop with various prompts, but that’s not why I am here today.
On average, my workflows take between 40-75 secs to complete. However, today I fired a complex query to the agent (it is a Postgres agent) and it took between 3-4 minutes. Just before the 2 mins mark the app returned a “Error communicating with AI agent. Load failed” error. I tried prompting Lovable to fix it but it kept saying it will increase the timeout to 5 mins, then to 10 mins, but neither worked!
Anyone with a similar experience that can help me cat past this? For complex data analysis you’d normally want the app to wait for as long as the agent takes to return meaningful business insights.
Thanks
Ashraf Dahab