EUS
Diagnóstico

Going back to work often means going back to the same small routines.

Opening several news sites.
Checking newsletters.
Visiting industry websites.
Looking for anything relevant that may have happened at a specific company.
Checking for new developments, investments, grants or market movements.

None of this is particularly complicated.

The problem is something else:

𝗧𝗛𝗘𝗦𝗘 𝗔𝗥𝗘 𝗧𝗔𝗦𝗞𝗦 𝗧𝗛𝗔𝗧 𝗥𝗘𝗣𝗘𝗔𝗧 𝗧𝗛𝗘𝗠𝗦𝗘𝗟𝗩𝗘𝗦 𝗘𝗩𝗘𝗥𝗬 𝗦𝗜𝗡𝗚𝗟𝗘 𝗗𝗔𝗬.

And that is precisely where a simple automation can have a much bigger impact than it may seem. ⚙️

Recently, I built a system in Make for a client with one very specific objective:

to receive only the information that was genuinely relevant to their business every morning, without having to search for it manually.

Nothing particularly futuristic.

No need to build a huge system just for the sake of it.

Simply identify a repetitive task and make it happen automatically.

🧭 The problem is not finding information

There is plenty of information out there.

Probably too much.

A company may need to monitor:

Doing it once does not take much effort.

Doing it every day of the year does.

But there is another problem.

If we automate the collection process and gather 40 articles every morning, we still have 40 articles to read.

That is why:

𝗔 𝗚𝗢𝗢𝗗 𝗔𝗨𝗧𝗢𝗠𝗔𝗧𝗜𝗢𝗡 𝗦𝗛𝗢𝗨𝗟𝗗 𝗡𝗢𝗧 𝗚𝗜𝗩𝗘 𝗬𝗢𝗨 𝗠𝗢𝗥𝗘 𝗜𝗡𝗙𝗢𝗥𝗠𝗔𝗧𝗜𝗢𝗡. 𝗜𝗧 𝗦𝗛𝗢𝗨𝗟𝗗 𝗥𝗘𝗠𝗢𝗩𝗘 𝗧𝗛𝗘 𝗡𝗢𝗜𝗦𝗘.

How the process works

The overall logic is quite simple:

SOURCES → UPDATES → FILTER → AI → RESULT

Let’s go through it step by step.

1. Choose what is worth monitoring

Before opening Make, there is a question that matters far more than any technical decision:

What information would actually be valuable to the company?

In the system I built, different RSS sources are combined:

📰 local and regional media;
📊 business and financial news;
🏢 information related to companies;
📍 sources linked to a specific territory.

The automation goes through those sources one by one and checks whether there is anything new.

But quantity is not what matters here.

𝗧𝗛𝗘 𝗤𝗨𝗔𝗟𝗜𝗧𝗬 𝗢𝗙 𝗧𝗛𝗘 𝗥𝗔𝗗𝗔𝗥 𝗗𝗘𝗣𝗘𝗡𝗗𝗦 𝗙𝗔𝗥 𝗠𝗢𝗥𝗘 𝗢𝗡 𝗧𝗛𝗘 𝗦𝗢𝗨𝗥𝗖𝗘𝗦 𝗧𝗛𝗔𝗡 𝗢𝗡 𝗛𝗢𝗪 𝗠𝗔𝗡𝗬 𝗬𝗢𝗨 𝗔𝗗𝗗.

Adding a hundred websites does not guarantee a better system.

Sometimes, it does exactly the opposite.

2. Make automatically checks what is new

Once the sources have been selected, Make checks the latest publications.

The system does not need to analyse the entire archive again and again.

It looks for new information.

📅 From Tuesday to Friday, it can focus on the most recent information.

📅 On Mondays, it can widen the period to include anything published over the weekend.

It may seem like a small detail, but it reflects something important:

an automation should adapt to the way the person using it actually works.

Not the other way around.

3. Collect the information you actually need

From each article, the system can automatically collect different pieces of information:

Make then brings all that content together so it can be processed as a whole.

At this point, we have already achieved something:

we no longer need to open every source manually.

But we still need to solve the most important part.

🧠 Deciding what deserves our attention.

4. Avoid receiving the same information again and again

This is one of those small details that separates a useful automation from an annoying one.

An article may remain in an RSS feed for several days.

Different media outlets may also report on exactly the same event.

That is why the system keeps a record of the URLs it has already used.

Before preparing a new radar, it checks that history.

After sending the new items, it stores their links again.

The logic becomes:

SEARCH → FILTER → SEND → REMEMBER → SEARCH AGAIN

The system does not just search for information.

It also remembers what it has already shown you.

And that greatly reduces duplicates.

5. This is where AI comes in

We could do something very simple.

Send all the articles we find to an AI and say:

“Summarise this for me.”

It would work.

But it would be a fairly mediocre automation.

In this case, the main role of AI is different:

𝗡𝗢𝗧 𝗧𝗢 𝗦𝗨𝗠𝗠𝗔𝗥𝗜𝗦𝗘 𝗘𝗩𝗘𝗥𝗬𝗧𝗛𝗜𝗡𝗚. 𝗧𝗢 𝗗𝗘𝗖𝗜𝗗𝗘 𝗪𝗛𝗔𝗧 𝗜𝗦 𝗪𝗢𝗥𝗧𝗛 𝗬𝗢𝗨𝗥 𝗧𝗜𝗠𝗘.

For example, we can ask it to detect signals related to:

📈 business growth;
🏭 investments or expansions;
👥 employment and recruitment;
🔄 corporate changes;
💶 grants;
📊 significant industry movements.

And discard things such as:

The difference is huge.

AI is not there to produce more content.

It is there to make sure we have less to read.

6. If there is nothing important today, it sends nothing

This is probably one of my favourite parts of the system.

The automation is not required to produce a report every day.

If it finds eight genuinely useful articles, it can work with eight.

If it finds two, it works with two.

And if it finds none…

it sends nothing.

Because:

𝗔𝗨𝗧𝗢𝗠𝗔𝗧𝗜𝗢𝗡 𝗜𝗦 𝗡𝗢𝗧 𝗔𝗕𝗢𝗨𝗧 𝗣𝗥𝗢𝗗𝗨𝗖𝗜𝗡𝗚 𝗠𝗢𝗥𝗘. 𝗜𝗧 𝗜𝗦 𝗔𝗕𝗢𝗨𝗧 𝗗𝗢𝗜𝗡𝗚 𝗢𝗡𝗟𝗬 𝗪𝗛𝗔𝗧 𝗡𝗘𝗘𝗗𝗦 𝗧𝗢 𝗕𝗘 𝗗𝗢𝗡𝗘.

The workflow itself includes a condition that stops the process when there is not enough useful information.

No filler reports.

No email just because “something had to be sent”.

7. 📩 The result: a radar, not a mountain of links

When there are relevant updates, the system transforms them into something much easier to consume.

For example:

Priority
What deserves attention first.

Headline
What happened.

Source and link
Where to find the full information.

Context
Why it may matter.

Quick interpretation
What it means in relation to the objective we have defined.

The result arrives directly by email.

And I think this is important too.

The person using the system:

They simply open their email.

The automation works around the person instead of forcing the person to work around the automation.

⏱️ How much time can something like this save?

Let’s imagine a fairly conservative scenario.

Every morning we spend:

5 minutes checking the press.
5 minutes checking newsletters.
5 minutes on specialist websites.
5 minutes checking specific companies.
10 minutes deciding what actually deserves our attention.

30 minutes a day.

It does not sound like much.

Until we calculate it over a full year.

With around 220 working days:

𝗠𝗢𝗥𝗘 𝗧𝗛𝗔𝗡 𝟭𝟬𝟬 𝗛𝗢𝗨𝗥𝗦 𝗔 𝗬𝗘𝗔𝗥.

For just one person.

And we are talking about a relatively simple automation.

The same structure can be used for many other things

The example I have explained uses news, but the concept is much broader.

🏢 Competitors

Detect when a company:

📑 Regulations

Monitor official sources and highlight only the changes related to specific topics.

💶 Grants and subsidies

Detect new calls and automatically apply the relevant criteria.

📋 Tenders

Find new opportunities by territory, activity, value or keywords.

🎯 Sales prospecting

Detect signals indicating that a company may be entering an interesting moment:

📊 Industry intelligence

Receive the genuinely important developments in a specific market every morning or every week.

The architecture is almost always the same:

SOURCES → AUTOMATION → CRITERIA → AI → RESULT.

Does this allow you to monitor the entire Internet?

Not exactly.

And I think it is worth being clear about that.

The automation in this example mainly uses RSS feeds because they are stable, structured and relatively easy to maintain.

When a website does not offer RSS, other possibilities can be explored:

But it depends on each website and the options it provides.

That is why I prefer to talk about:

automating the sources a company actually needs to monitor

rather than casually promising that we can “automate the entire Internet”.

AI is only one piece

There is something else I think is important.

This system uses AI.

But AI is only one part.

There is much more happening before and after it:

  1. Choose the sources.
  2. Check them automatically.
  3. Detect new information.
  4. Collect the data.
  5. Check what has already been sent.
  6. Remove duplicates.
  7. Apply criteria using AI.
  8. Decide whether there is enough value.
  9. Create the final result.
  10. Send it and store what was used.

That is what interests me when we talk about introducing artificial intelligence into a company.

Not putting AI on top of everything.

But asking ourselves:

𝗪𝗛𝗘𝗥𝗘 𝗖𝗔𝗡 𝗜𝗧 𝗥𝗘𝗠𝗢𝗩𝗘 𝗪𝗢𝗥𝗞 𝗧𝗛𝗔𝗧 𝗖𝗨𝗥𝗥𝗘𝗡𝗧𝗟𝗬 𝗔𝗗𝗗𝗦 𝗡𝗢 𝗩𝗔𝗟𝗨𝗘?

Start with small automations

When we think about AI and automation, we often imagine large projects.

But there is a much simpler way to start.

Ask yourself two questions:

What do we do almost every day in exactly the same way?

And then:

What part of that work is simply searching, copying, checking, organising or classifying information?

That is where some of the best opportunities often appear. ⚙️

Not because they are spectacular.

But because they repeat.

And a small task repeated hundreds of times can eventually become a very large one.

💬 Want the automation?

I have prepared a base version of this workflow so it can be adapted to other companies, industries and types of information.

𝗜𝗙 𝗬𝗢𝗨 𝗪𝗔𝗡𝗧 𝗧𝗛𝗘 𝗕𝗔𝗦𝗘 𝗠𝗔𝗞𝗘 𝗔𝗨𝗧𝗢𝗠𝗔𝗧𝗜𝗢𝗡

Write 𝗠𝗔𝗞𝗘 in the website chat and we’ll send it to you. 💬

From there, you can adapt it to the sources, industries and criteria you need.

A list of sources, a few filters, a little AI… and one less task to do every morning. ⚙️