@Nate Herk

@Nate "Community nodes" - Tavily - can only be used with self hosted n8n. Meaning cloud users cannot add community nodes. Nate still rocks, but the workflow is pretty much useless for Tavily nodes cloud-based n8n enthusiasts. Nate, OpenRouter charges extra for adding credits while promising a volume discount. Tavily community nodes aren't approved by default. Hope you enjoy using those..

@Roger Marquez Below is an end-to-end blueprint. It turns a few structured inputs (your business UVP + some basic ICP clues) into: 1. A rich, multi-section Ideal Customer Persona (ICP) document 2. 5–10 “click-magnet” content titles that match the UVP & ICP pains 3. Optional: long-form draft copy, short social snippets, and ad variants ► High-level architecture Trigger → Collect / look-up UVP & seed data → Clean & enrich → 🔀 Parallel AI tasks → Aggregate → Deliver (Start) └─▶ Webhook/Form trigger ├─▶ 1. “Fetch UVP & ICP seeds” │ └─▶ (Google Sheets, Airtable, or manual form) ├─▶ 2. “Normalize + trim whitespace” (Code/Function) ├─▶ 3. “Research enrichment” (SerpAPI, Clearbit … optional) └─▶ 4. “OpenAI – Persona builder” ├─▶ 5a. “OpenAI – Title generator” ├─▶ 5b. “OpenAI – Copy drafts” └─▶ 6. “Merge Results” (Merge node) └─▶ 7. “Deliver” ├─▶ Google Docs / Notion page └─▶ Slack / email (End)

@Nate Herk I've invested a whole $2,80 of Apify credits while making it work. Not counting OpeanAi and other. Whew, that was a journey :) I'm sure you might've heard about that already -- there's a challenge with output at Phase 4. And, Phase 5 too! Had to define using JSON Schema { "type": "object", "properties": { "doing_well": { "type": "array", "items": { "type": "string" } }, "dislikes": { "type": "array", "items": { "type": "string" } }, "wants": { "type": "array", "items": { "type": "string" } } }, "required": ["doing_well", "dislikes", "wants"], "additionalProperties": false } and update prompts with # Overview Respond ONLY with a valid JSON object matching EXACTLY this schema: { "doing_well": string[], "dislikes": string[], "wants": string[] } Do NOT add any other keys. Do NOT wrap it in triple back-ticks. Do NOT nest it inside "output" or any property. otherwise it wouldn't take doing_well, dislikes and wants and map it to a Google Sheet. I love the system!

@Ayush Lathwal Everyone will love you for a swarm or AI Agents, all linked to the legendary “MONEY” button (overnight delivery). The magical, mythical rectangle that deposits $500 into your bank account every time you push it...