I’m building a WhatsApp chatbot in n8n for a clinic. The bot needs FAQ handling, appointment booking, rescheduling/cancellation, reminders, and personalization I’m deciding between using a Switch node based workflow or the AI Agent node. My main goal is fast response time and a clean production design, but I also want the bot to feel personalized. For those who have built something similar: - Is the Switch/Rules-based approach better for speed and reliability? - Does the AI Agent node make personalization easier enough to justify the extra latency? - For clinic use cases, would you keep booking/rescheduling outside the agent and use it only for FAQ or fallback? - Any best practices for storing personalization data and routing intents in n8n? I’d like advice from people who have done this in production

Hi everyone, I’m building WhatsApp chatbots for my client to sell to doctors, and I’m trying to make sure the setup is structured in the right way from the start. The chatbot will work on WhatsApp, with n8n running in the background to handle the logic, doctor-specific knowledge, bookings, and alerts. These questions are very important for me right now: -> Should I clone the workflow for each doctor, or use a multi-tenant setup? -> Should I create different n8n accounts for each doctor? -> What is the safest way to structure this so it stays legal and compliant? I’d really appreciate advice from anyone who has already faced this kind of setup. Thanks in advance.

1. What I scheduled - I have 2 schedules running: 1) A daily morning triage where my calendar and emails get checked and reported on, and 2) An evening check-in to report on what got done, and what might be outstanding that needs to move to the following day and rescheduled. 2. Scheduled task or loop? I have these as scheduled tasks, because I want them recurring forever. I also have a task.md file that I created that gets updated daily and improved upon (I ask how we can do things better so it's always giving me ideas I may not have thought of). I'm also in the middle of doing a loop - I've just had it set to remind me every hour that the Google 1/O Keynote is on tonight (it's in my calendar anyway, but fun to play with the loop and know how I can use it for the future). 3. One surprise - it's wild to see the daily file being updated as we move through tasks. The productivity and accountability level is priceless. 4. Extra - I also created a dashboard. Once my task.md file is updated in the morning, I run it via my Chrome dashboard so I can visually see and move the tasks along via the boards to the next stage. It has a miniature donut chart at the top to show priorities. I have 5 different themes to choose from for viewing.

Today I've learned how to use MCP Server and get info I need from any website yo use the data later for different objectives. Here in the screenshots I've extracted Job Researchs only related to automation. At first The connection with Firecrawl didn't work, even following the steps of the vidéo but in the end claude fixed it on it's own. So good to also see that we can get any other informations from a website thanks to this !

Most law firms still route client inquiries manually. A paralegal reads every submission, figures out the case type, assigns it to the right attorney, writes a reply. Repeats this 20 times a day. I built a system that does all of that automatically — the moment a client hits submit. Here's what happens in the background: A client fills out a professional intake form hosted on a real website. The moment they submit, the workflow fires instantly. A JavaScript scoring engine evaluates their urgency level and referral source. It classifies them as Hot, Warm, or Cold in milliseconds. An AI Agent reads their case description in plain English and classifies it into the correct legal category — Personal Injury, Family Law, Criminal Defense, Business Law, Real Estate, or General Inquiry. The right attorney gets an instant Slack alert with the full case summary, confidence score, and client details. The client receives a personalized confirmation email — tone adjusted based on urgency. Hot clients feel like a senior attorney personally read their case. Cold clients get a professional standard acknowledgment. Everything is logged to Google Sheets automatically. The most interesting part — a divorce case involving a house dispute routes to Family Law, not Real Estate Law. Because the legal matter is the divorce. The AI understands legal theory, not just keywords. Tools: n8n · JavaScript · Groq (LLaMA 3.3) · Slack · Gmail · Google Sheets · GitHub Pages