No Code Backend Engine for Internal Link Automation [Part 2]

I finally set up something I’ve been wanting to build for a long time, a small backend engine that quietly runs in the background and keeps a record of all new blogs published and create semantic embeddings summary (to help my internal linking automation idea) 

It starts with something simple: the RSS feed.
Every time a new blog goes live, the RSS updates. And the moment that happens, the Pabbly flow wakes up.

From there, it does all the heavy lifting.

It pulls the new blog and extracts everything that matters

The flow takes the blog URL and breaks the page down into clean, usable parts:

• the title

• all H2 and H3 headings

• the cleaned body text
  (without images, footers, navigation bars, or any HTML junk)

Basically, it strips the post down to its pure meaning — the actual writing.

Then it creates a semantic summary for embeddings

This part is fun.

Using a custom prompt, the flow generates a short embedding summary that captures:

• what the blog is about

• the topics it covers

• the intent behind the content

• the structure reflected in its headings

This summary goes into the database and becomes the “embedding memory” of that blog post. Perfect for semantic matching, clustering, and — most importantly — internal linking.

Everything is stored in a clean backend database

Each new blog becomes a complete, structured record:

• URL

• title

• all headings

• cleaned content

• embedding summary

• timestamp

Over time, this creates a growing content intelligence layer behind the blog — something I’ll use for automations, recommendations, and even content revivals.

Why this matters

This system means:

• I never have to manually check new blogs

• I never have to copy-paste titles, headings, or text

• I never have to rebuild internal links after publishing

• I now have a database that “understands” my content semantically

It’s a small automation, but the impact is huge.
From the moment a blog goes live, it becomes part of a smarter, self-updating backend.

What comes next

Now that the backend is ready, the next layer is automatic internal linking.
Since every blog has:

• a summary

• a structure

• and a meaning stored semantically

The system can soon recommend:

• which older blogs to link

• where to link them

• what anchor text matches

• and how to keep everything updated as content grows

This is the foundation.
And honestly, it feels good to finally have it running.

Also Read: 
Part 1 of Internal Linking System [Backend Preparation]