ParentFull threadyz-yu·yes, now I found the time mainly caused by: 1. openAI API call, which could not be optimized for now 2. the time to build and deploy, which can be optimized by a preset setupView on HN