Everybody's talking about N8N, a free

and almost open- source platform for

building AI workflows and AI automations

using a graphical user interface. This

video today will be a crash course and

introduction into N8N. We're going to

learn how to use it by building and

implementing multiple workflows. We're

going to integrate Google Sheets, Google

Calendar, Telegram. We're going to use

AI agents and even Python code. This is

an interesting platform for programmers

but also for non- tech people. If you

like this video, let me know by hitting

a like button and subscribing. But now

let us get right into it.

[Music]

>> All, right,, so, we're, going, to, learn, how

to use N8N in this video today. Let's

address the elephant in the room right

away. I claimed that it's almost open

source. What does that mean? It

basically means the source code is

available. It is source available.

That's the correct term. But the license

is quite restrictive. You're not allowed

to modify or host it for someone else.

You're not allowed to build a competing

product. However, you can take the

source code, you can deploy it on your

own server for your company. You can

modify it. You're just not allowed to do

it for someone else. So, you can do all

of this for your company's internal

needs, but you're not allowed to build a

competing product, which makes it not

open source, but source available. So

running N8N locally is extremely simple.

All you need is you have to have node

and npm installed. And then you can just

open up a terminal and say npxn.

This is going to run n on your own

system. It's going to install it. It's

going to download everything that's

needed and then it's going to run the

software on port 5 678. Now the first

time you run this you're going to see a

registration form there. You have to

create an account for N8N. However, all

of this is happening locally in a SQLite

database I think. So you're not really

signing up for an actual service. In my

case here I have to sign in because I

already have the account here and I'm

just going to provide my credentials. By

doing that I can log into N8N. In this

case here, I already have some workflows

because I prepared them for the video.

But in your case, you probably have

nothing here. And you can just start by

creating a workflow. Now, before we

start working on an actual example, let

me briefly show you where you can find

what. The most important thing here is

the right side, this plus symbol or you

can also trigger it here. You always

start with a trigger. So the idea of N8N

is you have workflows. These workflows

have to start in some way. You can start

them manually by just clicking. You can

start them by form submission, by

chatting or by events from uh different

apps. For example, here on app event, I

can say from all these services I want

to wait for certain events and once they

happen, I want to trigger the workflow.

For example, a certain row could be

inserted into Google Sheets. I have here

on row added, on row updated, trigger

this workflow. Or I can also say on

PayPal event. So when something happens

there, I want to react to it. For the

sake of simplicity, we can start with a

simple trigger manually. Now, this will

trigger the workflow. Clicking on this

executes the workflow. And as you can

see, there's a connection here. So, to

this trigger, I can add now certain

things. So, for example, I can add some

AI action. I can add an action in an

app. And here again, we have many many

many different integrations. For

example, I can say go to Google Sheets

when I click and append a new row in a

sheet. Now for each individual node you

have different things that you need to

set up. Maybe a credential to connect to

Google Sheets, maybe some other API key.

And then you also need to say okay what

exactly happens which document, which

sheet, which row, what do you want to

do? But basically every note is

structured in the way that you get some

input and produce some output. So you

get the trigger without any data. So you

don't really have an input here, then

you do something, then you return

something, then you do something else.

You also have stuff to control the flow.

So you have if else, you have filters

you have uh transforms. So you can say

if something is the case, go here. If

something is not the case, go here. And

then you can go into AI and you can say

I want to have OpenAI and I want to

message a model with the data that I got

from the previous note. Now output that

data in a summarized way. Take that

again and feed it into Python code for

example. So I can go to core, I can say

code, I can say Python. And now take the

input, access it by using underscore

input, do some stuff here, output

something else. So all of N8N is based

on this idea of starting somewhere with

a trigger. Maybe the trigger already

produces some data. If it's not a click

but a chat message, for example, we can

take this message as input already, and

then every node takes this input, does

something, outputs something else, and

at some point the workflow ends. But

before we create a fancy workflow with a

lot of different nodes, let's start with

a simple one. We're going to start with

a trigger which is a form submission. So

we're going to click here on form

submission and here now the idea is to

specify what our form is going to look

like. So I'm going to say here this is

going to be a registration form and

we're going to say register

for beta. Then what we can do is we can

add form elements. So these are the

input fields. I'm going to say here add

form element name. This is going to be

text. I'm going to say placeholder your

name and this is a required field. Then

I'm going to add another one which I'm

going to call email. The element type is

going to be email. It's going to be your

email as a placeholder and that's also

going to be a required field. Finally

I'm also going to ask for the reason why

do, you, join?, It's, going to, be, a, text.

It's going to be your reason for joining

and it's not going to be a required

field. So once we have all of this, I

can execute this step. What this is

going to do here is it's going to listen

for a form submission. I can close this

now. And you didn't see this because I

have a tiling window manager in full

screen. But here it opened up a form. So

now I can say something like my name is

Florian, my email is whatever atmail.com

and my reason is because I want to. And

now when I submit this, this is

triggering this note. So we can take a

look at this note and we can see that it

pulled in a test form submission. And

now we have this output this JSON output

that is passed to the next note in this

case name Florian email whatever

mail.com reason because I want to and

then some information here about the

time and the form. So the next step is

take this and do whatever you want. So I

can take this and do some Python code

for example. Just let me show you so you

can see what this would look like. We're

going to use this later on. In this case

now I have these fields here and I can

use them in my code. What I do here is I

say for item in input all JSON new field

is one. So this would take the data

return it as it is but add a new field.

But I can also do something else. For

example, I can say item.json.name

is equal to item.json.name

lower. So that would make my name into

lowercase as you can see. So we can use

Python code to take in the input and

produce some other output. But that's

not what we're going to do here in this

example. I'm going to delete this again.

What we're going to do now is we're

going to take this data and append it to

an existing Google sheet. So, as you can

see up here, I have this N8 end sheet

this Google sheet, and it contains two

sheets. It has sheet one and beta

signups, which is what we're going to

use now. We're going to use the first

one later. And what we have here is

username, email, reason, and we have

Mike mike atmike.mmike and test. Now

what I want to happen is when I submit

the form I want a new row to be appended

to this table so that the data of the

signup is now in my Google sheet. In

terms of N8N this process is quite

simple but we still need to go through

the Google process as well. So we need

to go to console.cloud.google.com.

You need to log into your Google account

and then you want to create a new

project. Now in this case I already have

here an N8N project but I'm going to go

through the process with you from

scratch., So, I'm, going to, create, here, new

project and I'm going to call this N8N

tutorial project. I'm going to connect

it to my billing account. I'm going to

create this. Now once the project is

created,, I'm, going to, select, it., And

then we're going to go to the Google

Sheets API. So I'm going to click here.

I'm going to enable it. And now what we

need to do in order to use the Sheets

API in N8N actually is we need to set up

credentials. So in N8, if I create a new

note and I look for sheets and I say I

want to have a note that appends a row

in a sheet, it's going to ask me for

credentials to connect with. Now in my

case, I already have some because I

played around with it, but we're going

to create new ones now. So I have the

choice between OOTH 2, which is

recommended, and service account. So

we're going to go with O2. And for this

N8N provides me with a redirect URL. So

I'm going to copy this one. And down

here, we need to enter the client ID and

a client secret from Google Sheets or

from the Google console. So, I'm going

to go back here. I'm going to go to APIs

and services. If you're not here

already, you can go to the navigation

menu, APIs and services, and then

credentials. The first time you do that

you have to set up an OOTH consent

screen. So, you want to go here, and you

want to create a new screen. So, you're

going to go get started. App name is

going to be N8N app tutorial. Whatever.

I'm going to provide my email here as

support. I'm going to say external. I'm

going to, provide, my, email, again., Next., I

agree. Continue. Create. That's that.

Now, let's go to data access. Here we

need to provide scope. So, I'm going to

say add or remove scopes. I'm going to

look for sheets. And I'm going to just

enable everything that has to do with

Google Sheets. I'm going to save. And

then I think I need to also add myself

as a test user to be able to use that.

So, add users and my email again. Save.

And now we can go and create a client.

So I'm going to go to clients here and

I'm going to say create client.

Application type is web application.

Let's call this N8N client tutorial. And

here now we need to provide the redirect

URL., So, I'm, going to, paste, what, we

copied from N8N. I'm going to create and

as a result I get the client ID and the

client secret. Of course keep these

secret. You don't want to show them to

people. I'm going to censor this out or

delete the account. One of the two. And

here now we paste the client ID and down

there we paste the client secret. That's

basically it. The final action is to go

down and click this sign in with Google

button. This is going to open up a

window. It's going to ask me to log in

with my account. I'm going to pick the

one that is added as a test user. So

neural 9. And then I can go here to

continue. I can select all of these. I

can say continue. And now the connection

is successful. And I authenticated

myself here with the Google Sheets node

in NAD. So now we can close that and we

have to decide okay what exactly do we

want to do? We want to use a resource

which is a sheet within a document. We

want to append a row to that sheet and

the document that we're interested in

here. We can click on choose. Okay.

Actually we can see that we need to use

the Google Drive API as well. So let's

go back here. Let's go to drive. Let's

enable this API as well. And I'm not

entirely sure if we also need to provide

the permissions here. No. Okay, that's

enough. So, you just enable the Google

Drive API and then you select the sheet

that you're interested in. In our case

here, the N8 end sheet, which is maybe

not the best name. Maybe I should have

called it uh NAND document or whatever.

But in this document, we have sheets.

So, I can click on this as well. And I

can choose the beta signups. Now, what

I'm going to do is I'm going to map each

column manually. So, I'm going to say

that the username will get the value of

name. Now I don't need to do anything

fancy here. I don't need to type

anything. I can just drag and drop this

into here. Name is going to be username.

As you can see, I can also do it here

with JSON.name. I can also use

JavaScript functions to modify this. So

if you want to provide some JavaScript

code, you can do that here. But we're

just going to drag them in here. So

name, email, and reason. And that's

basically it. If I execute this now

it's going to take the information. It's

going to do that. And if I look into the

sheet, there you go. We have Florian

whatever mail.com because I want to. So

this means in my N8N workflow when I

submit this form, it's going to add a

new entry to the Google sheet. So let us

exit full screen here. Let's put this

into a separate window. Let's zoom out a

little bit. Zoom out a little bit here

as well. Now execute workflow. This is

going to open up this form. And now

let's go with test user. Testmail.est

test whatever and ABC. When I submit

this now, this is going to trigger the

workflow and we're going to add the row

to the Google sheet. Awesome. So, what

we want to do next is another

integration. I would like to get a

message on Telegram from a bot that

tells me that a new user signed up to

the list. So, for this here, I'm going

to create a new note. I'm going to say

Telegram and it's going to be send a

message here. So, send a text message.

Now once again we need to provide

credentials so that we can connect with

telegram. I already have it but I'm

going to do it here from scratch again.

Create new credential. And here we just

need an access token. So how do we get

this access token? Now here again I'm

censoring out my chats but what you want

to do is you want to search for

something called botfather. So botfather

is a bot that you can use to create

other bots. Then you can click on it and

it's going to open up the chat with the

botfather. Now the first time you use

this you're only going to get the info

message. You need to start a

conversation with a bot by clicking or

by typing /st start. Then it's going to

tell you what you can do. In my case

I've already created some bots here. But

essentially what you do is you say slash

newbot and this is allowing you to

create a new bot. So I provide this

command now. And now it tells me choose

a name for this bot. So let's say nadn

tutorial testbot. I hope this name is

not taken. There you go. Good. Now let's

choose a username for your bot. It must

end in bot. Now I'm going to go with the

same one. N8N tutorial testbot. Okay

it's already taken. So let's go with

testbotbot.

Actually I cannot edit the message. I

have to type it again. Testbotbot.

There you go. And since the name is not

taken, I can now use this. And what

happens is I get a token to access the

HTTP API. So I can just copy this token

and paste it to N8N. Then I can save

that. And this is a successfully created

credential. But that is not enough

because what I need to do now is I need

to instruct the bot who to send this

message to. And we're going to do that

by adding the bot to a chat. So here, as

you can see, I have resource message

send message and I need to provide a

chat ID. So I need to have this bot in a

chat with me and it needs to know that

it has to send this to the chat ID. So

the first thing we do is we click on the

link that gets us to the bot t.nutorial

/N8N tutorial testbotbot and now I have

to start a conversation with this bot.

So I click on start here as well and now

I have my bot in a conversation. It

doesn't do anything but what I want to

do next is I want to create a group and

add the bot to the group. So I click on

this button down here and I say new

group and then I have to add the bot N

tutorial testbot or testbot bot and then

I select it and I go and create the

group., So, now, we, have a, group, with, this

bot. Now the interesting thing is I

cannot just retrieve the chat ID. So

what I have to do is I have to add

another bot to the group to get the ID.

This bot is called the get IDs bot. So

you need to start a conversation with it

to have it in your contacts. And then

you want to go to the group you just

created. You want to click on it and you

want to add a member. So, I think my

camera is hiding it, but there's a

button down here which allows you to add

a member and I'm going to add here the

get ID spot. We add it. And now it tells

me here this is the ID of the chat. So

I'm going to copy that and I'm going to

go back to N8N and provide this as a

chat ID. And the text is going to be new

user signed up name is

username. I can add it here as well. And

email is whatever. And I can again do

whatever I want with that. I can use

some code here. But that should be what

happens. So if I execute this and

everything worked correctly, this is

executed and in the group I get the

message new user signed up. Name is test

user and email is test atmail.t test.

Cool. So let's try all of this again.

I'm going to exit full screen. I'm going

to close this. I'm going to close this.

And now I'm going to start with a form

submission. Maybe let me also add this

here as a separate window. Going to put

it down here. I'm going to hide the

sidebar., And, now, what, we're, going to, do

is I'm going to click here and I'm going

to say this is going to be John John at

johnh.jo

ABCDE EF submit. So what happens now?

New row is created in sheets. New

messages sent to Telegram. This is what

N8N is all about. Integrating multiple

tools and building workflows. If I may

for a second, I would like to plug

myself in as the sponsor of my own

video. If you go to my website

neural9.com, you will find a tab

services and a tab tutoring. Here you

can hire me for all sorts of stuff like

data science, machine learning, web

development. If you need help with

something in a project, here you can

book me for one-on-one tutoring if you

want me to teach you personally

something that you don't understand. If

you like my teaching style, on both

pages at the bottom you can contact me

via mail and also via LinkedIn. Just

wanted to let you know about this. So

let us now take a look at a different

kind of workflow. We're going to create

a different trigger which is going to be

a chat message sent within N8N. So the

idea is I can do test chat here. I can

say hello world and this is going to

produce an output in this note which is

a JSON object with chat input being

hello world. Now we can take this

message and feed it into a large

language model to get a response. For

example, I can go here AI open AAI

message a model. Now I'm going to use my

existing OpenAI account. All you need to

do if you want to connect your account

is create a new credential and provide

the API key. For this, you would go to

platform.openai.com/api

keys, create new secret key, and then

you would copy it. But as I said, I

already have this. So, I'm going to

choose to message a model. I'm going to

choose the model GPT5. So, I'm going to

scroll down here, GPT5.

And the prompt to this model is going to

be the chat input from the previous

note. So, if I execute this step, it's

going to take hello world, feed it into

GPT5, and produce this output here.

Hello, how can I help you today? Now

we're going to do something more

interesting. What we're going to do is

we're going to take the message and

we're going to create a calendar entry

based on the message. So, the large

language model will have to create a

structured output, a JSON object that

represents starting time, end time

title, and description for an event and

then it's going to be added to Google

Calendar. However, the adding is not

going to happen via tool use. So, it's

not going to be the large language model

that creates our calendar event. We're

just going to create a structured JSON

object and then we're going to create

the event with N8N. So to be precise

the message is going to look like this.

Based on the user message, figure out

what the calendar entry should look like

and return a JSON object with start

end, title, and description. And then

we're just going to drag and drop the

chat input down here. That is going to

be our prompt. We're going to scroll

down as well and say output content as

JSON. And when I execute this now, I

mean, in terms of hello world, we're not

going to get a very good output, but we

can then formulate a message like I have

a barber appointment at 4 p.m. tomorrow.

So, as you can see, in this case, it's

just title, hello world. Description

hello world. Doesn't make a lot of

sense. So, let's go back and try to

provide a different message. Tomorrow at

400 p.m., I have an appointment at my

barber. So, this took a bit more time

already, but you can see it says 23rd of

October. Today is 22nd. That's correct.

And it provided also the time here and

barber appointment at my barber. Now, we

want to take this JSON object and create

a Google calendar entry. So, I'm going

to click on the plus here. I'm going to

go and look for calendar. We can see

Google calendar here. And I'm going to

choose create an event. Now, the

credentials here are different from the

sheets credentials. So, we need to

actually create new credentials for this

one with the same key and the same ID is

fine, but we do need to change some

things in the Google console. So, we're

going to, go, to, the, Google, console., We're

going to go to the Google calendar API.

You want to enable it. In my case, I

already did this here. You will find a

button here again, enable. And then we

need to go back to the OOTH consent

screen. So, APIs and services, OOTH

consent screen. And in here, we need to

add scopes for data access. So, we're

going to go to data access, add or

remove scopes. We're going to look for

calendar. We're going to select Google

calendar API. And even though it might

be overkill, I'm just going to give all

the access here. Update. Scroll down.

Save. And in case you already lost your

client ID and client secret, you can go

back to clients. You can go to N8 client

tutorial or whatever you called it. And

down here, you can add a new secret. So

in this case, I just have to copy here

using this button. I'm going to go and

create a new credential here. Client

secret. The ID stays the same. It's up

here.

And that's basically it. Sign in with

Google again. Select your account.

Continue. Continue. And there you go. In

your case, you probably have to select

the check boxes. I already went through

the process once. Cool. So once we have

this here we need to instruct it to

create a new event in a calendar. For

this first of all create or select a

calendar in my case social@ner

neurallonline.com and now we can provide

start and end date. So let's remove this

and let's also remove this because our

start is going to be the start provided

by GPT and our end is going to also be

the end provided by GPT. Then the title

and description are additional fields.

So I'm going to go down to summary. This

is the title. I'm going to drag and drop

the title here and then the description

is going to be the description. Cool.

So, I have my calendar opened up here. I

have nothing in here. And now, if I say

execute step, this should take an action

in my calendar. I actually don't see it.

So, let me refresh this. And there you

go. Barber appointment. Now, the time

you will notice is a bit off because of

time zones. We could also instruct GPT

to provide a different time zone. We

could also do it manually here with

JavaScript. I'm just going to leave it

like this for the sake of simplicity

but it added this with the proper title

and also with a proper description. Now

finally, we can also make N8 respond to

the chat. So, I can go to this chat note

here and go to add field response mode.

I can say using response notes. Then I

can close this. I can create here a

response note respond to chat and I can

say successfully created event 4 and

then whatever the summary was dot and

then I can save that and now we can go

through the whole workflow again today

at 5:00 p.m. I have wrestling class and

there you go successfully created event

for wrestling class and in between we

had our model output this we had our

calendar create the event there you go

wrestling class again with the problem

of the time zone but this is how it

basically works okay so now let us go

and build something a little bit more

comprehensive I have a Google sheet here

this was sheet one of my N8 end sheet

where I have some sales that are

documented so I have the date the sold

item the quantity and the unit price.

Now, I want to build a workflow that

takes this data, aggregates it, compares

my sales today to my sales yesterday

and if the sales today were lower or

less than yesterday, I want to get a

large language model to explain to me

why and send it via Telegram as a report

and alert. And if I earned more today

compared to yesterday, then I want to

get a Telegram message that

congratulates me on that. So this is not

necessarily the most complex task, but

it will require a lot of notes because

we're not just going to do all of this

in Python. So we're going to start with

a manual trigger. I want to do this on

demand. And the first thing we want to

do is we want to pull some data from the

Google sheet. So Google Sheets and then

we want to get rows in sheet. We're

going to use the credentials we already

created. I'm going to go with sheet

within document. Get rows. The document

is going to be again N8 sheet. The sheet

this time is going to be sheet one. And

for options, we're going to say output

formatting is going to be the values

unforatted. So execute step, this is

going to give me all of the data as JSON

object like this. Next step is I want to

filter. I'm only interested in today and

yesterday. So I'm going to say that the

date of the item has to be today or

yesterday. Now for this, we're actually

going to use a JavaScript expression. So

I'm going to say here new date date.now

now and then minus basically one day to

iso string slice 010. So it's just going

to take the date not the time stamp

which is basically yesterday and it has

to be greater than that. So it has to be

after or equal to. So now when I execute

this it's only going to give me the

sales of today and yesterday. Next up I

want all of these prices here to be

combined into total prices. So quantity

and price I want them to be multiplied

so we get a total price. I can do that

easily with Python. So we'll go here to

core run code Python. And now I'm going

to say for all items JSON.tal is going

to be equal to item.json

dot quantity with capital Q times item

JSON dot or actually not dot because we

have a space in there. So I'm going to

use this syntax unit price and this now

gives me a total for every single item.

Next step is to aggregate. Now the

aggregate node is actually not really an

aggregate node that is a bit confusing I

think and actually it's in data

transformation because in order to

aggregate in terms of how we aggregate

in pandas or something you actually use

the summarize note as you can see here

sum count max across items is the

summary note and here I basically just

say sum over total and we split by date

so then I get the total today and

yesterday with the date here let me

restructure this a bit to save some

space. The next thing we're going to do

is another Python node where we're going

to now take the JSON object and turn it

into today and yesterday instead of this

structure. So for this let's go again

core, code, Python., And, now, I'm, going to

copy paste this one to not waste your

time. But basically all we're doing here

is we're importing daytime sdt. We're

saying today is this date. Yesterday is

this date. Then we also have the strings

and basically we just say if the date is

today then return the object today with

the sum total of this day otherwise

return yesterday with a sum total of

that day. So execute this and we get

yesterday and today which now makes it

very simple to take this and perform an

if, statement, on, it., So, we're, going to, go

with flow if and we're going to say if

today is less than

yesterday and now we're going to have

two branches. One is going to fire. In

this case today is less than yesterday.

So we go to this true state here.

Otherwise we're going to go to the other

one. Now actually let us handle the

false one first because if we have a

success we just want to send a telegram

message. So, I'm going to say here, send

a text message to this account. The chat

ID is the same as before. So, in my

case, this here. And we're going to say

great, you sold more today. Or let's say

you earned more today compared to

yesterday. And this should send a

message in Telegram. Now, here things

get a little bit tricky because what we

want to do if we sold less today

compared to yesterday is we want to

analyze the situation. But if I go into

this note here, you can see I only get

yesterday and today. I don't get any

context information. I don't get any

transactions. I don't get any sales. I

just get the sum. So what we need to do

if I want to analyze all of this is I

also need to somehow get back the

initial information. For this we're

going to use another Python node. I'm

going to go to code here again. Python.

I'm going to copy paste this here again.

What we do is we access data that we had

access to before. So as you can see this

code in Python beta 1 is the last one

but we also have this code in Python

beta without the one. And here we have

still the sales information. So what

we're going to do is we're going to say

node items is equal to that specific

note. So code in Python beta the first

one dotall.2ython

then I have the sales. I get the JSON

object for the individual items. Then I

list them and then I add all of this as

sales to the JSON object. So when I

execute this what I get here is I get

yesterday today and then I get the list

of sales with all the items and what

happened. Now with this output I can now

just attach another open AI note here

messaging model. Let us go this time

with GPT4. The analysis is not going to

be very complicated. And here I'm going

to provide the following prompt. Let me

maybe make more space so we can read it.

Essentially it's going to say look at

these sales. Then it's going to take the

sales, parse them as a JSON string

summarize why we earned less today with

the sum of today compared to yesterday

with the sum of yesterday. Keep your

answer very concise. Just describe the

important pattern and formulate it like

a message to a user in a chat. Why?

Because we're going to send it to

Telegram. So if I take all this and

execute this, you will see that it tells

me you earned less today because only

one laptop was sold while yesterday had

higher volume, three laptops and five

monitors sold. That's great. So finally

we're going to take that and send it to

Telegram. So Telegram again, send a text

message. Chat ID is same as before. And

this is now going to be the message.

Execute step. And in Telegram I get

here, you earned less today because only

one laptop was sold while yesterday had

higher volume, three laptops and five

monitors sold. So if I go back to my

sheet now and I change this to 10 and I

run the entire workflow on demand

what's going to happen? Let's click

here. Execute workflow. All of this is

run. And we get here. So now I get the

message, great, you earned more today

compared to yesterday. If I turn this

back to two, let's say, save this.

execute workflow. All of this is going

to happen. A message is going to be sent

and I get the analysis again with why I

earned less. Now, finally, let me

briefly show you how we can use AI

agents for this. Let's go with a chat

trigger. Again, we're going to chat with

a model. And what I'm going to do is I'm

going to, add, not, just, an, AI,, but, an, AI

agent. So, an AI agent can have multiple

things attached to it. One is I can add

a model. For example, let's go with an

open AAI model again. I'm going to go

with GPT40. Then we can also add some

memory. For example, simple memory in

N8N, which allows us to keep track of

what was already talked about. And then

we can add tool use. For example, we can

add stuff like other workflows that we

already have. For example, call N8N

workflow tool. I can now say go and

execute this workflow 3 that we just

had, which was the one for comparing

sales. So I can click on this and now

whenever it thinks that this has to be

triggered it will trigger it. We can

also say when it should be triggered

when the user wants to compare sales of

today versus yesterday. Call this

workflow. So now I have it stored here

and we can start chatting with the

model. Test chat. What are my sales

today compared to yesterday? Now I think

I actually have to do it outside of

testing. So let me send this again. This

is going to trigger the agent. Okay

actually 40 cannot use tools. So let's

switch to 4.1 mini. That should be fine.

And let's try again. And there you go.

It recognizes it should use that

workflow. It executes that workflow and

I get a notification on Telegram that I

earned less today. And again the

reasoning. So yeah, of course you can

take this then as well. I also got the

answer here as a result, but we can also

continue and do something else with it.

So that's basically how you can work

with N8N. There's much more to explore.

many different integrations with many

different ways to authenticate yourself.

But you can also take a look here at

templates and see how other people are

working with N8N. You have workflows

that people have created that you can

just copy and adjust if you want to.

Let's go with this one for example, use

for free. I can just click on it. I can

say import template to localhost. And

now I have it on my system. Here it

instructs me to first do some stuff. I

can skip that and just take a look at

this template also with some tutorial

and description here. But this is quite

complex and you can see that it collects

idea and image. Then it creates an image

with nanobano which is an image

generation model. Then it generates a

video ad script. It generates the video

itself. And then also it posts it onto

all platforms. This is quite advanced

but you can also explore that if you

want to. You can click on the different

notes and see exactly what they're

doing. This is how you can also learn

how to use NAN. So that's it for this

video today. I hope you enjoyed it and

hope you learned something. If so, let

me know by hitting a like button and

leaving a comment in the comment section

down below. Also, if you're interested

in personal tutoring or services, check

out my website. I have a services and a

tutoring tab there where you can book me

or hire me for freelancing services or

one-on-one tutoring. And of course

besides that, don't forget to subscribe

to the channel and hit the notification

bell to not miss a single future video

for free. Other than that, thank you

much for watching. See you in the next

video and bye.