We have successfully updated the vLLM engine in the production environment. This deployment fixes the recent structured output errors; consequently, AI Insights is now fully functional, and all automated workflows with structured output nodes are operating as intended.

We have successfully identified the root cause of the current slowdown. Our team is currently validating a fix in our staging environment.

You may experience increased latency when generating AI Insights during this period. We appreciate your patience.

AI Insights and Automations may experience temporary reliability issues due to intermittent GPT API inaccuracies.