My current n8n workflow turns a product feed into Google Ads image creatives through a mix of prompt-engineered GPT-4 calls, a Stable Diffusion endpoint, and the Google Ads API. It runs, but the conversion lift I expected never showed up. Here’s what I need over the next 1–3 days: • A rapid audit of the existing scenario (JSON will be shared once a short NDA is in place). • Smarter prompt chains that generate stronger, on-brand visuals and copy variations. • Tighter automation logic—error handling, rate-limit strategies, and leaner API calls—so each creative lands in the correct ad group every time. • A measurable bump in conversion rates after a live test flight (I’ll supply the baseline data). You’ll likely touch n8n (or, if you prefer, Zapier/Make) as well as OpenAI, Stable Diffusion, and the Google Ads API. Please come in ready to push quick commits, explain each change, and hop on fast feedback loops in chat—this is a short sprint, not a long project. I ignore generic pitches; a concise note showing a similar AI automation you’ve built will get my attention immediately.