data collection#
n8n can collect data as you want with webscraper or RSS. Let’s say you chose the easy way: RSS. You received URL data via RSS. n8n can make HTML requests to the corresponding page with its Web Request component. In fact, with HTML Extract, it can take the page source, locate the content holder from within and get the content instantly.
You can either clean this data and create training data for yourself, or you can rewrite it according to the rules you define. rewrite? How will this happen? Yes, your path has reached the door of LLM. Now you can easily get the token and continue with any LLM you use. Additionally, n8n also documents how to do this for you. You got the API token and got started. With a nice user-system prompt… great? no it’s not. Because the API pricing of the LLM you are using is different from the pricing you have on the web and on the client. Your monthly payment does not cover the API. -Bad news :(- APIs generally work with a ‘pay-as-you-go’ logic. So what should you do? You should use a cheap but definitely stable LLM, but how? This is where great solutions like GROQ come in handy. It has the same stability as your LLM but hosts multiple models much cheaper and doesn’t send you a bill you’ll regret at the end of the month.
Data Processing#
Now the material is ready. But what about the image/images – if the content you are going to create will be so boring and dull that it is just plain text, then I have nothing to say, but the internet user decides what something is like by looking at its cover first. Isn’t there a solution like GROQ for this? Actually, there is. There is a great site called REPLICATE. Here, all the image making models and their prices are written so far. Moreover, using its API is child’s play. But one thing is very important: price/performance. Because the data you provide will extract an image prompt from your content, and from there you will go to REPLICATE with the prompt you have. But for which model? This is where pricing comes in. To get a quality attractive, realistic, even creative image solution, you will have to spend some money. Otherwise, the images don’t turn out to be very attractive. (You can choose flex/schnell to be cheaper, but there is not much difference between creating an image with flex/schnell and having a kindergarten student create it :D)
It’s not really possible to enter without calculating the cost, is it? So will you publish this data as Internet content? Where? On your own site. What about its social media marketing? X, LinkedIn, Reddit, HackerNews, BlueSky… which one? With what parameters? When? Yes, now you are also working on social media marketing. Honestly, this is the most challenging step. (Like the boss at the end of the chapter) Because it’s not important that you send your content to these social media platforms. It is also important that you comply with their publishing principles. Otherwise, you may be banned. In fact, you often need to do it manually. Or you can train an actual AI to do it. I have not seen any such project yet. Social Media Marketing Editor AI 🙂 Really a good idea.
Last words… For those who think AI is zero cost or very low cost, the experience ends with disappointment. Like bosses who misunderstand artificial intelligence because they pursue zero or very low costs.
<a href