In this video, I break down four real client projects we sold this year, what each agent did, why it was valuable, and exactly how I priced and closed them. You’ll also get a 7-step roadmap you can follow to start selling your own AI agents, avoid the traps I hit early on, and position yourself as a strategic partner instead of just a freelancer.

I just dropped a new video where I share the story of how I sold my very first n8n automation back in 2024. What might shock you is how simple the system actually was. Looking back, I could probably rebuild it in 30 minutes today. But that’s the point, clients weren’t paying for a “complex workflow,” they were paying for the outcome. This was the first time I made money selling an automation, and it taught me something that’s even more true now: business owners don’t care how advanced the system looks, they care if it saves them time, money, or focus. And today, people are even more open to investing in solutions like this, as long as you can communicate value clearly and show you’re someone they can trust. Curious to hear your thoughts, does this change how you think about selling automations or agents?

I have created an n8n workflow that solves model selection, routing, and data persistence via a unified API (i.e. no more individual integrations or manual adjustments when switching models). In short: the foundation for the European "Open Router" in n8n is therefore laid... 🧩 What the workflow does: * Connects n8n with Requesty via an OpenAI-compatible interface * Supports both fixed model usage and dynamic selection * Stores selected models in Google Sheets * Uses the chosen model for response processing * Technically clean separation of selection, control, and execution ⚙️ Two setup variants: 1. Directly via OpenAI node: * Base URL: https://router.requesty.ai/v1 * Model selectable directly in the node * Easy to configure 2. Dynamically with HTTP workflow: * Model list is retrieved via API * Selection in the form trigger (dropdown) * Storage in Google Sheets * Automated further processing with the selected model 🛡️ GDPR-compliant with Requesty: * Unified API for multiple models * Provider with DPA and based in the European legal area * No direct communication with US APIs required * Setup reduces legal complexity and technical fragmentation Follow-up questions or ideas for further development are, as always, welcome 🤝 PS: Attached is the workflow. It will also appear on n8n in a few days, provided the submission is accepted...

I built this and submitted it as a free template on the n8n Creator Hub. While I'm waiting for review and approval, here’s a clear overview of what this intelligent LinkedIn comment system can do: 🌍 Multilingual and context-aware - The system automatically detects the language of any LinkedIn post. - It generates replies in the same language. - The output sounds natural and relevant to the context. 🎯 Adapts to tone and style - It analyzes the sentiment of the original post. - The comment style is adjusted accordingly – from professional to casual. 💡 Smart and cost-efficient setup - Prompting is designed to keep model usage lean and affordable. - Claude Haiku 3.5 and GPT-4.0-mini are fully sufficient for high-quality results. 🔗 Fully integrated with LinkedIn workflows - The system automatically mentions the original post author. - A matching LinkedIn reaction is set based on post content and mood. 🛡️ You stay in control - Manual approval happens via Telegram before anything is published. - You decide what goes live and when. ✅ Ideal for: - Professionals who want to engage authentically on LinkedIn. - Teams or companies with limited social media bandwidth. - Anyone looking to scale high-quality engagement without scaling effort. Status update:I’m intentionally waiting for n8n’s approval before sharing the link. Quality assurance is important to me.The live test has already proven successful. Once the template is officially published, I will share the link here. 📸 I’ve attached screenshots of both the generated comment and the full workflow so you can get a clear picture of how it works. What do you think? Are you using LinkedIn automation already? Who would like early access to test this out? Cheers and thank you for being my involuntary test candidate, @Nate Herk 🤝😉

Hi everyone, I'm working on a workflow that automatically processes transcripts from Microsoft Teams. The goal is to extract structured outcomes from meetings; things like: action items, decisions, and task lists. The current setup is a working prototype. Whether I turn it into a production-ready solution and publish a tutorial depends on how much interest there is. The workflow is built with n8n and fully customizable. If this sounds interesting, check out the LinkedIn post below and let me know in the comments which extended features you’d find useful. Also feel free to share if you have clients who could benefit from this kind of automation (maybe even without realizing it yet…). https://www.linkedin.com/posts/stefan-z-23a07b262_aitools-meetingproductivity-gdpr-activity-7346465419211292672-W8Rg?utm_source=share&utm_medium=member_desktop&rcm=ACoAAEBwIo4BZWnyZQ1uW7ZuYY8xDOuR9cjwBbM If there’s enough feedback, I’ll finalize the workflow and write a clear step-by-step guide. Beginner-friendly and easy to implement. 🤝