Lite OSINT: NotebookLLM, Open-Source RAG?

finx

3 min read

·

Jun 18, 2026

--

Listen

Share

At first glance, NotebookLLM is a learning site provided by Google.

But if examined the site features properly: NotebookLLM is an open-source RAG.

Sounds absurd?

Let me prove my case with possible use case for OSINT:

Justification as of why NotebookLLM is an RAG

Before I show, let’s circle back on the definition of RAG:

RAG or Retrieval-Augmented Generation (RAG) is basically a way to make AI instead of referring their old training dataset, we give external sources like (database or documents) for them to add with AI capabilities.

Press enter or click to view image in full size

In update v5(Update v5 (Malaysian OSINT resource list)- AI stuff | by finx | May, 2026 | Medium), we established some ground rules for the RAG that we want to use:

1) Use the select sum of sites to gather the data

2) Able to be shared globally

3) Able to withhold heavy usage

4) Less setup overhead on my end (no payment preferably)

Okay, lets walk through each rules and examine if NotebookLLM works shall we?

1. Use the select sum of sites to gather the data

Press enter or click to view image in full size

It is established that each notebook able to have 50 sources, so: for our use case, it is good

2. Able to be shared globally

3. Able to withhold heavy use

As long as you don’t breach the free usage limit tier on the screenshot, I think we’re good.

4. Less setup overhead on my end (no payment preferably)

As seen: you can get the NotebookLLM only via your Google email, this is quite possibly the lowest bar to enter.

So, NotebookLLM is that unicorn that I presented on Update v5. It is right under my nose all along😅

Shall we walk through how to use it then?

How to use NotebookLLM for OSINT

1. Gather your sources, for this POC: I would use the staff directory list I have in Update v4 (Update v4 (Malaysian OSINT resource list) — PSE and Dorking | by finx | Medium)

[Uni directory list]
https://fssm.umt.edu.my/chemical-sciences-program-lecturer-2/
https://www.uim.edu.my/v4/about-uim-2/structure/directory
https://portal.utem.edu.my/oas/directory/stafsearch_SMSM.asp
https://cyberjaya.edu.my/university/staff/staff-directory
https://aviation.edu.my/staff-directory/
https://registrar.ums.edu.my/staffdirectory/web/public/index
https://www.iium.edu.my/directory/
https://library.oum.edu.my/oumlib/staff_directory
https://www.ukm.my/pusatsiswazah/staff-directory-2/
https://mitrans.uitm.edu.my/index.php/directory
https://hep.utem.edu.my/direktori-staf/
https://pppl.umt.edu.my/directori-staf/
https://konvokesyen.uitm.edu.my/v2/index.php/direktori
https://telefon.uthm.edu.my/
https://www.um.edu.my/list-staff.php?kodPTJ=GF&kodJAB=GF01
https://directory.usm.my/
https://direktori.unimap.edu.my/DIREKTORI/index.jsp?ABC=
https://www.utm.my/directory/
https://uitm.edu.my/en/discover-uitm/administration/office-directory
https://miat.unikl.edu.my/staff-directory/

2. Copy all your sources in NotebookLLM. There are two ways to do so:

3. Let the model read the sites and start your query

Press enter or click to view image in full size

And, that’s pretty much it. You got a powerful notebook for you to do recon on your target or whatever you want to utilize on

With this, I end this blog,

Thank you for reading