n8n

There are moments that remind you why you build a community. Today is one of them. A year ago, we hosted the very first n8n Fest. Today, more than 500 community members + the n8n team came together again at SAGE Berlin. To everyone who joined us: thank you. From across the city to across the world, you helped make today what it is, special. A LOT more photos and videos are coming! 🙂 If you are at the Fest, share yours!

Remember our AI agent dev tools report? No worries if not. This blog post catches you up on the three things that matter most in agent development right now: ▪️ Identity: what an agent is allowed to do, and on whose behalf ▪️ Reliable execution: agents that finish what they start ▪️ Intent analysis: what the agent was trying to do, not just what it did Written by Andrew Green. Read it NOW. 🙂 https://lnkd.in/daJwqnUH

Connecting one MCP tool is easy. Running MCP in production is where the real engineering starts. Our latest guide covers: - Why production MCP is about authentication, observability, and deployment, not just protocol support. - The three MCP integration patterns teams use in production. - What to look for in an MCP-compatible automation platform. - How n8n works as both an MCP server and an MCP client, so AI assistants can use your workflows while your workflows can connect to external MCP servers. Read it here:

How is Mercedes-Benz scaling AI across their organization? 🚗🤖 During the second edition of n8n Business Lab, held at IPAI in Heilbronn, Germany, Michael Deptuch and Moritz Strenger from Mercedes-Benz AG's AI Center of Competence walked us through their AI Ecosystem strategy. Their approach is built on three layers, each designed for a different kind of team and use case: ▪️Integrated AI: off-the-shelf tools ▪️Composable AI: low/no-code workflows ▪️Build-Your-Own-AI: custom solutions for expert teams with specific needs n8n powers the Composable AI layer. It gives teams across the business the ability to build AI agents and automate workflows, all while integrating cleanly with existing internal systems. Watch the full keynote now live on our YouTube channel 👇 https://lnkd.in/dmrrvfXD

Tell us who's actually building the automations where you work. 🙂

Choosing a cloud integration platform is a big commitment, and getting it wrong is expensive. A platform that doesn't scale or keeps your data locked in a black box means growing costs and technical debt. And the right choice isn't about which tool lists the most integrations. It's about the deployment model fitting your security needs and having enough visibility to troubleshoot when things break. We put together a comparison of several platforms to help you find the architecture that works for your stack without locking you in. 🔗 Full comparison here: https://lnkd.in/eDEHGpQg

Upload a photo of a dress. Get back an image of a model wearing it and a video of them walking in it. All generated by AI. This week's featured template connects VLM Run's virtual try on API with Telegram, Discord, and YouTube. You upload a clothing image, the workflow pairs it with a model photo, generates a realistic try on image, creates a fashion walk video, and publishes everything automatically to the platforms you choose. If you're in fashion, e-commerce, or content creation, this is the kind of workflow that turns a single product photo into ready to post content across multiple channels. 🔗 Check it out here: https://lnkd.in/gTcNAFq6

Not every AI workflow failure should be handled the same way. A timeout, malformed JSON, a hallucinated answer, and a failed API call all look like "failures" – but each needs a different recovery strategy. Blind retries often add cost without fixing the underlying problem. Our latest guide shows how to build layered fallback logic for AI-powered automation, covering retries, validation, fallback models, human approval, and dead-letter queues in n8n. Read the full guide: