Hey all, I just built a complete AI call center that can handle any process - customer support, sales calls, lead qualification, appointment booking, order status checks, refunds, and even creates support tickets. In this video, I walk you through exactly how to build this using Ultravox, N8N, and Twilio. This system can be customized for any industry - real estate, e-commerce, coaching, law firms, you name it. It's saving businesses hundreds of thousands of dollars while driving massive revenue increases. You can get the free template and all the prompts in the classroom! Go check it out here - https://youtu.be/medCBnEV50I Drop a comment and let me know what industry you'd use this for and what you thought about the video!

Hi Guys, i build an Call Agent with Ultravox. It works very well. But in case, the Agent, can´t answer the question, he should transfer the call to an employee. So the Employee get a Call over Twilio and insert the generated conference room. Now i will transfer the active Call (Twilio status: in progress) with POST https://api.twilio.com/2010-04-01/Accounts/{{json.body.twilioSID}}/Calls/{{ JSON.parse(Object.values($('Get Call Data').first().json)[0]).providerCallSid }}.json With the Body: Twiml <Response> <Say voice="alice" language="de-DE">Sie werden zu einem Mitarbeiter weitergeleitet. Bitte bleiben Sie dran.</Say> <Dial> <Conference startConferenceOnEnter="true" endConferenceOnExit="true" {{ $('Extract Call Data').item.json.conferenceName }} </Conference> </Dial> </Response> I got "The resource you are requesting could not be found" 404 Response. But the call is active in twilio. Doe´s anyone have an idea?

I want to know the costs. I want to work as a call center employee in artificial intelligence via N8n. How do I create it? How much does it cost per month

What is your guys' favorite llm?

Hey everyone – wanted to share some thoughts and practical experience about using MCP (Multi-Channel Prompting)with n8n to build an agent system that can automatically create new agents just from a simple text description of what you want them to do. 🤖 What’s the big idea? Imagine writing something like: “Create an agent that listens for new customer requests in Gmail and adds them to Airtable as CRM records.” …and the system auto-generates the whole n8n workflow, credentials, and nodes – no manual setup required. That’s exactly what you can build with n8n + MCP: a “factory” workflow that reads a human description and creates the n8n automation flow that matches it. ⚡️ How does it work in practice? Here’s a step-by-step approach you can implement: 1️⃣ User Input - The user types a natural-language description in a form or chat interface. - That text is sent to an LLM node (like OpenAI, Ollama, etc.) in n8n. 2️⃣ Prompt Engineering - The LLM node has a carefully crafted prompt that tells it to generate valid n8n JSON for a workflow. - For example: 3️⃣ Create Flow via n8n API - The generated JSON is posted to your n8n instance via its REST API. - This automatically creates a new production-ready workflow without manual clicks. 4️⃣ Feedback / Notification - The system notifies the user: ✅ Real Example I built an experimental setup where the user just types: “Read new Gmail emails with subject ‘order’ and save to Airtable.” The LLM responds with n8n JSON defining: - Gmail Trigger - Filter Node - Airtable Create Record This JSON is then posted via API to the n8n instance – and the workflow is live in under a minute. 🌱 Why is this so good? - Removes barriers for non-technical users. - Super fast scaling – from 1 to 100 agents in minutes. - Low maintenance – update prompts centrally. - Teams can build their own workflows with plain language. It’s a beautiful fusion of no-code automation and AI-native development. 🚀 The Future of This