Powered by n8n and GPT. Writes, scores, and fixes ad copy on autopilot.
Love this
This is amazing! Would you share the n8n template?
Just sent you a LinkedIn invite. Once we’re connected, I’ll pass the template along!
Hi John, I'd also love to see the template, as I'm technical but adding more marketing to my daily activities. Would that also be possible?
Yes of course, do drop me a connect on Linkedin.
Would love to put this to the test. Could you share the template with me?
Hey Veena, drop me a LinkedIn connection and I'd be happy to send it over!
Done :)
Excellent work .. Thanks for sharing. Can you consider sharing n8n JSON too.. Thanks!
Sure! Let's connect over Linkedin.
This is amazing. Really cool to validate and build off of it. 1. I've run ~25 through and not one has scored a 9 or above. Is this normal? 2. Any way to repeat the loop over and over until it scores a 9 or above :-).
Perhaps try strengthening the prompt inside your evaluation node? Good idea on the looping though, will explore it :)
Care to expand on using json prompts? Are you parsing them into text for the agent or simply pasting the json in system prompt
Yes that's right. I am parsing them (each copy framework) individually into the prompt for the agent.
Love this
This is amazing! Would you share the n8n template?
Just sent you a LinkedIn invite. Once we’re connected, I’ll pass the template along!
Hi John, I'd also love to see the template, as I'm technical but adding more marketing to my daily activities. Would that also be possible?
Yes of course, do drop me a connect on Linkedin.
Would love to put this to the test. Could you share the template with me?
Hey Veena, drop me a LinkedIn connection and I'd be happy to send it over!
Done :)
Excellent work .. Thanks for sharing. Can you consider sharing n8n JSON too.. Thanks!
Sure! Let's connect over Linkedin.
This is amazing. Really cool to validate and build off of it. 1. I've run ~25 through and not one has scored a 9 or above. Is this normal? 2. Any way to repeat the loop over and over until it scores a 9 or above :-).
Perhaps try strengthening the prompt inside your evaluation node? Good idea on the looping though, will explore it :)
Care to expand on using json prompts? Are you parsing them into text for the agent or simply pasting the json in system prompt
Yes that's right. I am parsing them (each copy framework) individually into the prompt for the agent.