Speaker A
Hey there, everybody. In today's video, I'm going to be walking you through this 100% automated blogging system so that you can generate content at scale for yourself or clients. This workflow makes it super easy to input some information
Learn how to create a 100% automated blogging system using AI agents in n8n with human-in-the-loop control and SEO optimization.
Key Takeaways
- Automated blogging workflows can scale content creation while maintaining quality through human oversight.
- Combining multiple AI tools and platforms enhances flexibility and customization in content generation.
- Human-in-the-loop design improves control and relevance of AI-generated blog posts.
- SEO optimization is essential for making AI-generated content discoverable.
- Sharing and managing AI agents via platforms like Lerty enables client collaboration and monetization.
Summary
- The video demonstrates a fully automated blogging workflow using n8n, Lerty, and OpenRouter AI models.
- Users can input company details, brand voice, and industry keywords to customize blog generation.
- The system includes an AI assistant for brainstorming blog ideas and human-in-the-loop approval steps.
- Lerty serves as a communication layer to interact with AI agents and share workflows with clients.
- OpenRouter allows flexible use of multiple AI models via a single API key for content generation.
- The workflow pauses for human input on blog topic, audience, goal, CTA, and word count before generating content.
- The AI performs research, outlines, and writes blog sections with progress indicators visible in Lerty.
- Human control is emphasized to balance automation with oversight and customization.
- The system supports SEO optimization to improve blog searchability.
- The workflow can be integrated with platforms like WordPress or Wix for publishing.
Chapters
- 00:00Introduction to Automated Blogging System
- 00:56Overview of Tools: n8n, Lerty, OpenRouter
- 01:50Configuring Company and Brand Details
- 02:39Using the AI Assistant for Brainstorming
- 03:43Human-in-the-Loop Blog Topic Setup
- 04:42Setting Blog Goals, Audience, and CTA
- 05:50Generating the Blog Post with AI
- 07:52Workflow Progress and Section Mapping
- 09:00Content Generation and Quality Assurance
- 11:29Integration and Final Thoughts
Full Transcript — Download SRT & Markdown
Speaker A
about what you want the blogs generated about, and then it just runs for you.
Speaker A
This workflow has human approval, so you can approve certain steps. It has an assistant you can talk to, so you don't just have to generate blogs, but you can actually brainstorm ideas. And it has so much more involved, like an SEO optimizer
Speaker A
that will actually go through and keyword optimize based on your blog post in order to help you get searched. So, if you want this template, we have a free resource down in the description breaking down this entire workflow, and
Speaker A
you can basically just copy and paste it and get using it. We use three tools for this. We're using N8N, which is what you're seeing right now. We're also using Lerty, which is our tool that we developed, a place where you can
Speaker A
connect, communicate, and share your agents to other people. So, if you wanted to share this with a client and charge them for it, Lerty would allow you to do that, or you can just have it in your own dashboard, which is
Speaker A
kind of nice. And then for some of the models, we are also using OpenRouter.
Speaker A
OpenRouter is just a place where you can input one API key and select from pretty much any model imaginable, which makes it great for tasks like these because you can test different models against certain things. But without
Speaker A
further ado, let me show you how this works. So we have two parts of this workflow. We have an assistant on the top line where you can brainstorm ideas or your client can brainstorm ideas about what blog posts to generate with
Speaker A
this assistant because this assistant in the instructions knows everything about the company, which you can input, and it also knows everything about the workflow. So it knows which way to properly guide you in order to generate a really good blog post. So this
Speaker A
assistant can be talked to and queried at any time. We also have this end-user config node. So you can just come in here, and if you were onboarding a client or you're trying to set this up for yourself, you can easily just put in
Speaker A
details about the company that are going to be fixed. So every agent in the workflow knows exactly what to do, and they know exactly what industry they're talking to. They know exactly everything about what you're trying to do with
Speaker A
these blogs. So we have company description, company name, brand voice. So how do you want these blogs to feel?
Speaker A
How do you want them communicated? You can edit that here. And also industry. So what keywords are we going to be focusing on? What types of blog posts are we going to be generating? That type of thing. And then we have this other
Speaker A
route. And this other route starts with a human in the loop. So it waits for the human to say, "Okay, this is what I want to generate a blog about." Let me show you an example so you can see this in
Speaker A
action. And then I'll get into explaining all of the other nodes. So like I said, you can talk to this assistant just as you would any other agent. I'm going to be using Lerty as my communication layer. I'll leave a link
Speaker A
down in the description. You can actually sign up for a free Lerty account and just follow along for free during this entire process. But what I'm going to do is I'm going to just talk to my agent since it's connected with a webhook right over here. So it can see all the chats I have coming in. And I'm just going to message the assistant. I'm just going to say, "What can you do?" And now we can have a conversation with this AI
Speaker A
agent just like we would any other chatbot. It says, "I brainstorm fresh blog ideas with you." And we can say, okay, research the latest trends in AI and generate me some ideas. And right now, I kind of like having a
Speaker A
brainstorming process with generating blog posts. I don't really like having it be 1,000% automated running in the background. I kind of like this thing I call automated control. I like when everything is automated, like the drafting of ideas,
Speaker A
the drafting of posts, but I also like having control in the situation, which is why we add human in the loop in these ways to brainstorm different ideas with your agents. And so, I'm just going to send this off and let's get some blog
Speaker A
ideas. This is great if a client or an operations layer within a business is using this or yourself because the ideas that this blogging agent can come up with are recent. They're industry specific, and it's just some things that
Speaker A
people are actually searching for. So I like option number one. And so now that we have blog ideas, what we have is a command that starts this entire blogging process. You can make this command anything you want. You don't even have
Speaker A
to have it go to the assistant. You could have run this right out of the gate, no matter what the user says. But that command right now is slash start.
Speaker A
Once we type in slash start, what happens is it pops up with a human in the loop node where you actually get to type out the blog topic. You get to put the audience segment, the goal, you have an
Speaker A
optional CTA field. So if you're trying to sell something, you're trying to sign people up for a newsletter, getting them to buy your product, you can do that here, getting them to just go back to your website. And you also can get a
Speaker A
rough word count whether you want a 500-word blog post or a 5,000-word blog post. And right now in the workflow, as you can see, it says waiting for a response. So the N8N workflow is pausing right now. This is a human in the loop
Speaker A
node. So it's waiting for the human to respond. If I go back here, this red section is the one where it's waiting, this interactive message. So once we fill out the details necessary, then this entire workflow process is going to
Speaker A
run based on those details, which makes it super customizable and just a very good user experience, especially for integrating this into a company and enhancing a certain operation. This can be done with anything. So with that in mind, let me go back and actually fill
Speaker A
out this form. So, like I said, I liked that topic right here. And so, I can just copy that and paste it in topic.
Speaker A
And for audience segment, maybe I just want to say beginners with AI and automation who want to get into no-code tools like N8N. Nice goal. So, this is where we need to start thinking about like what's the overarching goal
Speaker A
of making this blog post. This is important because this is going to give your agent something to chase and something to go after when actually creating the content. So, it's not just creating aimless content that doesn't matter. Well, I just said for a goal,
Speaker A
help people understand the process of automating single tasks at a time in order to enhance portions of their workflows rather than replace them. I think that's a very important concept, and so I think for a goal that'll be a
Speaker A
great start. In CTA, let's just say I want people by the end to have signed up for my newsletter, the AI Roundup. Maybe I have a newsletter called the AI Roundup.
Speaker A
Rough word count. We can make it 2,000 words. So quite a lengthy blog post.
Speaker A
Nonetheless, I'm sure it will be great. So then it's as easy as just hitting generate blog. And like I said, you could automate all of this and just have AI cycle through this automation, and you can set it up to where it's automated,
Speaker A
but I like having a little bit of control and letting AI do the heavy lifting, but me setting the topic and orchestrating the model. So now what I'm going to do is hit generate blog. And now it's going to get cooking. So right
Speaker A
now it's studying. So that part of the workflow is currently on the researching step. So with Lerty, we add those custom typing indicators so we can indicate how far along the workflow is within the actual interface right here.
Speaker A
So I don't have to keep going back and looking at the execution logs. I just know now that I
Speaker A
So I don't have to keep going back and looking at the execution logs. I just know now that it's at the researching stage. When it gets to the outlining stage, we will see a new typing indicator appear, which is kind of
Speaker A
another benefit of Lerty. Like I said, you can sign up for a free account. Link in description. Makes the whole process more fun, more seamless. But you can do this with really anything. If you have your own telegram system or email
Speaker A
system, you could also implement it in that. But as you can see now, it says outlining. So what's happening right now is the blog post is currently going through the section mapper. And what the section mapper is doing is it is
Speaker A
planning out basically the entire blog post. It's taking in all of the company context. It has all of the inputs. So the things like the word count, the topic, the audience segment, the goal, the things that we input on that form in
Speaker A
specific on this run, it has all of that that it's taking into account and then it's trying to map out sections for this blog post. And a member of our team actually built this automation and it is insane how good this section mapper is
Speaker A
working. As you can see, each section word count must be between 250 and 750 words. It's very specific, not too short, not too long, but just enough to get in the information you need. And this is being optimized for readers and
Speaker A
retention. So, keeping people on the blog post, getting them to click through, getting them to be intrigued as they're reading, and then it's outputting the sections in a very specific way for the research loop to take place, which is down here, the blog
Speaker A
writer. Let's see what part of the workflow it's at right now. So, as you can see, it's actually generating the blog post, and it's on section four of eight. We have a nice little section tracker here which keeps track of what
Speaker A
stage of the blog post we are currently at and that's just within that blog writer loop. Now keep in mind after the section mapper there was that keyword researcher that finds the very high targeted keywords that's going to
Speaker A
surface this blog post on Google starting section five of eight and this will be done in no time. All right, it's starting its last section, section 8 of 8. And then it will format the blog post and then generate its result within the
Speaker A
chat. And just like that, we have the blog post generated. So, it's the soloreneur trap. Manual work eating your time and profits. Picture this. You're hustling as a soloreneur. Coffee in hand, staring an inbox piled high with leads. You sip through emails, copy
Speaker A
paste data into spreadsheets, hunt for client info online, draft social post. So, it goes on doing all of the pain.
Speaker A
And then it hits it with this line, leaving little energy for the big picture growth that builds real wealth.
Speaker A
So, it's actually capturing what happens to solopreneurs very well. And then it goes through and it just created this entire blog post for us in a matter of minutes. And then it breaks down what AI agents are and actually a very good way
Speaker A
with the brain, the memory, and the tools that is very similar to how I would actually describe it, right? The the models in the AI agent are the brain of the workflow. They have memory which keeps track of past interactions and
Speaker A
then you can give agents tools so they can use real apps. Yeah. So this blog post is actually using real stuff and really nice sources and then it just keeps going. I mean this thing is packed with keywords. This thing is ready to
Speaker A
go. It even gives uh links where people can go in order to sign up for things.
Speaker A
And as you can see I went to that link and yeah, it did actually work. So it's not just like fabricating information.
Speaker A
It's actually finding sources and using those sources to create this blog post. Gives a local install guide and it just keeps on going. I mean, this is a 2,000word blog post and we can even put that to the test. And so, what I'll do
Speaker A
right now is I'm going to copy this and see how close to 2,000 words it got.
Speaker A
Head over to Docs, paste it in, and then I'm going to go to tools, word count, and we are just shy of 2,000 words. So, it does a really good job at getting very close to the actual word count,
Speaker A
which is unlike any other regular chatbot. if you just try to do that. And so the beautiful thing is we now have this blog post that we can even go further with. So what just happened was it just sent that blog post to the
Speaker A
agent. But if we wanted to, we could have another human in the loop after this that says send to WordPress, send to Squarespace, Wix, send to anything.
Speaker A
It doesn't really matter. Or we could even add a little revise button where we could select revise and then select what we would actually want to revise. Let me show you what both of those options might look like. But now I'm just trying
Speaker A
to show you like the flexibility of this workflow and what you could do with something like this. If you want to make it more automated, if you want to add more control, if you want to send it to
Speaker A
certain people, send it to certain platforms, revise it, archive it, you can do so usually with human in the loop. That's why human in the loop in 2026 is so important and why companies are actually hiring their employees back
Speaker A
after firing them in 2024 and 2025. People big companies are starting to realize the importance of having a human in the middle. And so that's why it's important to add steps in your workflow where you can make decisions and it
Speaker A
makes things break a lot less and it has a lot less fail rate when you actually add that properly. And so the user gets options and you can add as many options as you'd like. You know, this is very
Speaker A
nice, very versatile. And each one of these options can be a different route in NADN. So, if you hook it up to something like a switch node, you could have, you know, seven different options based on what button they click over
Speaker A
here in Lerty. And you can use the data from the workflow along with going down a different route of the workflow. Let me take this from the top and show you what this might look like in action. I
Speaker A
won't hook up the switch node right now. I just want to show you how this surfaces after the message gets sent in the chat. So, I'm just going to run through it again real quick. And so, now what I'm going to do is I'm going to
Speaker A
generate another blog post in order to show you this new human in the loop feature. I'll keep it off camera for now just so you don't have to watch the entire thing once again, but I just want to show you what those look like when
Speaker A
the options appear. And furthermore, when you select one, what the success payload looks like in N8. So, I've got all my information filled out. Now, I'm just going to hit generate blog. All right, so I just generated a 500word
Speaker A
blog post. It came back in a nice and beautiful format, just like always. We have the blog post just going to work.
Speaker A
But what happens next? Well, as you can see, you can add in these custom payloads in order to customize your workflow even further. For example, I could hit send to WordPress, archive, or revise. Each one of these buttons could
Speaker A
take a different action with this content. So, let's just say I didn't like it. I could hit revise, and now that response was submitted. So if I go back to the execution logs, what I want you to notice is so if I go back to the
Speaker A
execution logs in NAN and I copy the most recent execution to the editor, what you're going to see is if I double click on the last node, we have a response type buttons and then what selection did I make? So the selection I
Speaker A
made was titled option because I forgot to title it different in the payload. If I go back here, you can see that since I selected revise, it output the word option. But the point is is I could change that to revise and then this two
Speaker A
would say revise when selected. And what we can do with that is we could add something in here like I said like a switch node and we can say if selection is equal to revise then go down the revise path. And so in
Speaker A
that case it was but since I named it wrong I'll have to do option for now.
Speaker A
But then when I select it, it goes down that revise path. So if I have more paths in here, maybe one for send to WordPress. I can rename that the WordPress output. And I go out here. As you can see, the revise route is green
Speaker A
because I selected the revise button. And so now I could go on and do like hundreds of different things with this workflow. It's super scalable for any use case. Like I said, I have a free resource in the description where you
Speaker A
can download and implement this workflow and get it up and running super fast. You can come in here and you can edit the system instructions to be uh unique for what you want it to be. If you enjoyed, please like and subscribe. I
Speaker A
would highly appreciate it. And drop a comment letting me know what automations and agents do you want built out. I'm going to be looking at the comments and building them as we go and giving you these free resources here on YouTube.
Speaker A
So, make sure you subscribe if you want more and make sure that you comment letting me know which ones you actually want to see so that I can build them out. Furthermore, if you want to join a community of AI enthusiasts and
Speaker A
learners, people who are building the future with artificial intelligence, then I recommend joining our AI foundations community. I will leave the link in the description below. We have calls every week. We have hackathons. We have big threads. We're probably one of
Speaker A
the quickest groups to meaningful AI news. We have everything you need in order to scale up your skills in this new everchanging environment.
Topics:automated bloggingAI agentsn8n workflowLerty platformOpenRouterhuman in the loopcontent automationSEO optimizationno-code toolsAI content generation
