Hi !new to all this and obviously i have questions regarding the WF image attached.The WF is aiming to generate seo for ETSY from images , the first GPT node correctly extracts detailed metadata from my artwork images (subjects, style, medium, mood, themes, etc.). However, when I pass this metadata into the second GPT node to generate Etsy SEO (titles, tags), the output is often generic, repetitive, and inaccurate,irrelevant
Probably is due to different categories of the given images, landscape to illustration etc,any idea how to fix this?
(Select the nodes on your canvas and use the keyboard shortcuts CMD+C/CTRL+C and CMD+V/CTRL+V to copy and paste the workflow.)
This is what I managed to achieve. You can analyze my nodes and the changes onto yours. Please note I had to change values in expressions in all nodes, to the data structure output from Gemini which will be different than yours. If you don’t have Gemini, register and get an API key for free.