Hey everyone, I’ve just organised the resource hub into a single classroom section with folders for each category of resource. I hope this makes it easier for you to find the templates and content you’re looking for! Cheers, Brendan

Learn to build your own AI Graphic Designer using N8N & Nanobanana. It’s capable of automatically generating LinkedIn carousel slides from a single message as well as generating social media hiring posts. These are a couple of common use-cases we’ve needed within our agency, and now, I can simply ask our agent via Telegram exactly what I need, and it’s able to generate the designs in seconds. The benefit here is that this AI Agent is trained on our agency & our design styles. Meaning we’re not just getting any AI-generated slop that looks like any other AI images. With the new Nanobanana model by Google, we can control how much of the image we want to edit, so this system can utilize existing designs to work with. Check it out 👇

Anthropic just dropped a massive post on how they built their multi-agent research system. So, here's how n8n users can steal these ideas to make their workflow automations 10x better: Anthropic's key insight? When they asked an AI to find all board members of every tech company in the S&P 500: - single agent: completely failed, got stuck in sequential search hell - multi-agent system: crushed it by splitting work between specialized agents This is the playbook hiding in plain sight for n8n builders: 1. Stop building one massive agent workflow that tries to do everything. that's like hiring one employee to run your entire company. 2. Create a "lead agent" in n8n that breaks down complex tasks and delegates to specialized "worker agents" — just set up a main workflow that triggers sub-workflows. 3. The magic happens when these agents work in parallel. A task that would take 2 hours sequentially gets done in 10 minutes. 4. Anthropic found their multi-agent system outperformed single agents by 90.2% on research tasks. That's not a small improvement — it's the difference between success and failure. 5. n8n makes this easy for non-coders: AI agent node → break task into chunks → call specialized sub-agent. 6. Bonus tip: give each worker agent only the tools it needs. One might use Google Search, another might use Gmail, a third might use Notion. Keeps things clean. 7. in n8n, use the little star button to let the AI figure out parameter values on its own (to, subject, message, etc). Massive time saver. 8. Anthropic discovered something fascinating: agent systems burn through tokens 15x faster than chat interactions. But for complex tasks, it's worth every penny. 9. Look at your current n8n workflows — which ones fail because they try to do too much at once? Those are perfect candidates for the multi-agent approach. 10. Stop treating your n8n AI agents like solo artists. they're meant to be in a band, each playing their part of the symphony.

We’ve just hit 10,000 members in the community! A huge thank you to everyone who’s joined and been part of the journey so far. I’m committed to keeping it active, engaging, and packed with value for all of you. Let me know below if there’s anything you’d love to see added to the community!

Vapi has just released a native SMS integration with their agents, enabling texts to be sent from your agents and texts to be sent to your agents, receiving messages live on the call. In this video, I will do a complete step-by-step walkthrough of setting up a Vapi AI Voice Agent and enabling real-time communication with SMS.