The Weekend Business - A SaaS for linking blog posts with AI

This is the first post in a series where I'll share business ideas that excite me, along with the technical challenges they'd require to become reality. The goal is to spark interesting discussions about both the potential for growth and the obstacles we might face when trying to turn an idea into a successful venture. Some of these ideas could be the seeds of future companies, while others are just fun thought experiments. Either way, I'm excited to share them with you and hear your thoughts on how they can become a reality.

The Problem

I love reading and writing blogs. In order to get attention and make your blog post popular, you need to do some advertising. I’m not an expert, but I do know that backlinks and quality links, in general, give you a huge boost. There are thousands of blog posts on how to get backlinks, but this idea isn't about asking for or buying them. Instead, it's about creating backlinks from the blog posts you’ve already written over time.

Example

• Post 1 (about healthy breakfast ideas): "To learn more about the benefits of a balanced diet, check out my detailed post on nutrition basics."
• Post 2 (about nutrition basics): "For some quick and healthy meal ideas, don't miss my guide to breakfast recipes."

Platforms like Medium often recommend similar blog posts, but sometimes these recommendations are random and not very related. These recommendations usually appear at the end of an article, but what if related topics could pop up while you're reading? This way, readers could dig deeper into a topic and then return to the original blog post, keeping them engaged and informed.

However, if you have many blog posts, it becomes challenging to: 1. Find the perfect location to link another related blog post. 2. Find the right blog post without getting lost in a jungle of articles you've written.

As a result, most people simply link to the posts they remember or take a wild guess and search for related content. This is suboptimal because these backlinks are valuable both for reader engagement and for SEO.

You end up: - Losing engagement from your user. - Losing screen time (because if they go to another post and come back, that's more time on your site). - Losing traffic due to lower ranking, which could improve if you linked more intelligently, especially on platforms like Medium or even subreddits.

The Solution

An app that scans all your blog posts and makes recommendations for cross-links between them.

User Journey:

1. The user enters their profile URL.
2. The app scans all their blog posts.
3. It shows a table of blog posts that could be cross-linked.
4. Optional (for premium users):
   • Suggested links to specific paragraphs for cross-linking.
   • Automatically adds an indentation with a link at the end of the paragraph, confirmed by the user.

Technical Blueprint

• The user enters their profile name (e.g., airabbitX.medium.com).
• The app fetches a list of all blog posts from this profile.
• It sends the titles to an LLM like GPT and asks which titles seem related.
• It reads the recommended pairs and asks the LLM for paragraph-level links.

Proof of Concept

Here's an example of how this could work. I copied my profile page into GPT:

Which of these stories might be related and could make sense to cross-link? Create a table.

How to safely delete a Hugging Face model from the cache I have a large collection of machine learning models on my device and stored in the Hugging Face cache. Many of them I no longer need or... Published about 1 hour ago 2 min read

Chatting with Your Private LLM Model Using Ollama and Open Web UI We've already gone over the first two options in previous posts. In this post, we'll look at how to chat with your private model using... Published about 7 hours ago 3 min read

Zero-Prompt Engineering --- How to integrate LLMs at the FS level Large Language Models (LLMs) have transformed AI development, making it easier than ever to integrate them into applications. Yet, despite... Published on Sep 15, 2024 4 min read

And here's what I've got, which all make perfect sense. 

Interested in the code for the mockup? Just drop me a PM, and I'll be happy to share it with you!

Simple Mockup

I don’t want to spoil too much and will let you make up your own mind. I'm sure you have even better ideas than I do.

But to demonstrate the concept, here’s a simple mockup of how this could look.

The user enters their profile and gets a link recommendation. They can then dive into any of those recommendations and get more refined suggestions on a paragraph level.

Here are some paragraph suggestion links I received from ChatGPT:

• Linkify
• Crosslinker
• Recommendr
• Connecto
• Linkmaster
• Interlink
• Recommender
• Linkbuilder
• Crosspost
• Linkmatrix
• Linkgraph
• Linkmap
• Linktree
• Linkhive
• Linkforest
• Linkweb

Conclusion

That’s basically it. If you feel motivated to implement this idea, go ahead! If you know someone who's already working on this, we’d love to hear about it and see how others have solved this problem. For those who implement the idea and are part of our community, we’re happy to share your app (and solution) with everyone here. Stay tuned and happy hacking!