Hello again! 👋
Following up on the migration to n8n I mentioned in the previous post, today I bring you a practical implementation that is already running on this very website: the new contact form automation.
And it's not just a simple connection to receive an email. I took advantage of code nodes and AI integration to create a genuine security shield. 🛡️🤖
The Problem: "Smart" Spam
Anyone with a contact form open to the public knows it: spam is a plague. And I'm not only talking about the typical bot that fills in fields at random. I'm talking about messages that look real but try to sell dubious SEO services, scams, or simply make noise.
Captchas help, but they often bother the legitimate user and don't filter the content of the message itself.
The Solution: n8n + Semantic Analysis with AI
Thanks to my new n8n server, I designed a workflow that receives each form request and subjects it to rigorous analysis before bothering me with a notification.
How Does the Flow Work?
- 01Webhook reception: The form sends the data (name, email, message) to my secure webhook at
n8n.broslunas.com. - 02AI analysis: Here's the magic. Instead of simple rules, I pass the message content to an Artificial Intelligence model.
The AI analyzes the context and intent of the message. It detects whether it is an unsolicited commercial proposal, a phishing attempt, or a legitimate message from a user or client. * It doesn't rely only on keywords, it "understands" what is being said.
- 01Filtering and action:
🔴 Spam/Malicious: If the AI determines it's junk, the process stops there. I don't even get a notice. The system learns and blocks. 🟢 Legitimate: If the message is valid, the flow continues. It is nicely formatted and a priority notification is sent to my private channels.
Immediate Advantages
Since I activated this system yesterday, I have reduced the spam reaching my inbox to ZERO, without losing a single important message.
In addition, since it's my own system hosted on my server (self-hosted):
- +Total privacy: My users' messages don't pass through unnecessary third parties.
- +Absolute control: I can adjust the AI's "sensitivity" or add new specific rules whenever I want.
- +Speed: All of this happens in milliseconds.
This is just the first step of what I plan to do with n8n. The ability to integrate complex logic and Artificial Intelligence into my own automation flows is opening a world of possibilities for me.
What do you think? Would you trust AI to filter your important messages?
Until the next automation! 🚀