I Replaced n8n With Claude Code (AI Agents Got 10x Easier)

So instead of building out another workflow inside tools like naden zap here or make.com we're going to be building out our first agentic workflow using clawed code in this tutorial today. Before we actually go into

today. Before we actually go into showing an example of what this looks like and building it out I want to tell you the two main reasons why you should ever consider even using a tool like

cloud code. The first one has to do with

cloud code. The first one has to do with time. You're building out workflows to

time. You're building out workflows to earn you money, save you time or for some objective. Ideally, you'd want to

some objective. Ideally, you'd want to spend the least amount of time actually building those workflows out. The

problem with tools like NAN is that you actually have to spend weeks, months, or even years learning the tool, understanding every single node, understanding the logic behind it, connecting it all together, debugging it, all of that kind of stuff. But with

cloud code, you're just messaging it like you'd message chatbt or claude, telling it, hey, I want you to build out a workflow. It will go ahead build it

a workflow. It will go ahead build it out. It'll debug everything and it'll

out. It'll debug everything and it'll run it for you. That's number one.

Number two is that there are features inside cloud code that were not even accessible in tools like naden. Let me

show you by explaining the example that we're building out today. Now in today's example, we are going to be automating probably the most boring task in the history of humanity, which is year-end

bookkeeping and accounting. And the

reason I'm smiling is because I don't even think accountants and bookkeepers enjoy the process of doing this. And I

know that if you don't like the idea of automating this in like 10 or 20 minutes, then you probably absolutely hate the idea of manually doing this in 20 to 40 hours at the year end. So

here's how it works. We're going to be taking invoices, which are expenses and revenue streams. And then we're going to be dumping that into a tool like for example Air Table. We can upload this to

QuickBooks as well or Zero, whatever your accounting software is. And then we can actually build out a custom dashboard to model all of this data like the revenue, expenses, net profit or tax collected. So we know how much we

collected. So we know how much we actually have to pay at the end of the year. We can have a profit and loss um

year. We can have a profit and loss um chart as well. And we can even break down expenses by categories and then ask questions to Claude like, "Hey, how can I cut back on expenses to save money or

whatever else you want to ask Claude?"

In addition to that, we can actually take these invoices and then we can stamp processed or whatever word you want on every single one of them. And

inside cloud code, we can take all of our PDF files. We can sort them automatically based on folders of each month. And then we can rename them as

month. And then we can rename them as well. So historically they were service

well. So historically they were service provider name followed by the date and now we're reversing that. So it's the date followed by the service provider's name. You can also get a report at the

name. You can also get a report at the end of the day with all of our revenue and expenses, our top clients, our top categories, all of that kind of stuff.

And it's really not going to take very long to do this. And inside a tool like NAD, well, a lot of that is just not possible. Like we can't build out a

possible. Like we can't build out a dashboard inside NAD. And we also are very limited in how we can actually ask

NAD to pull back information. So we can store data in air table using a tool like naden but actually retrieving information back from it to ask questions like hey what are our top 10

clients how much revenue did we generate all these kind of things are very limited in nature and for that reason cloud code is just a better tool overall it's faster more functionality and it

gets us better results okay so let's go into building this out just before we do I want to mention cost here cloud code does cost money specifically you will need to be on the pro plan in order to access this and that's $17 a month.

We're also going to be using Claude API for this which is going to cost a couple cents as well. Now, in order to get started here, we have to head over to Google and type in Cloud Code. Hit the

first link and of course like good oldfashioned Claude is giving us really complicated instructions to actually go ahead and set this up. Talking about

things like PowerShell and Windows CMD and all this kind of stuff. Um, and if you're a non-programmer, this might look really confusing to you. Let's break

down the two simplest options. We need

to use something called a code editor. A

code editor just simply looks like this.

You essentially have files on one side.

You can look at the files inside of this. And then we can have cloud code on

this. And then we can have cloud code on the right hand side where we can just ask it to do everything for us. Now

there's two popular choices. We can use something like Visual Studio Code. Okay,

this is free. And we can also use anti-gravity. And this is free as well.

anti-gravity. And this is free as well.

The choice is yours. You can use both.

I'm going to be using anti-gravity for this tutorial. just before you download

this tutorial. just before you download either or you do need to download NodeJS. This is the underlying code that

NodeJS. This is the underlying code that both of these run off of. Okay, so with anti-gravity, we'll head over to their site and we are going to download the desktop application. Where um cloud code

desktop application. Where um cloud code comes into the picture is that it's actually an extension inside of anti-gravity or VS code. So, I'm going to open up a new window here. And you

can think about like anti-gravity as the code editor and then cloud code as an extension that sits on top of anti-gravity for us to interact with.

So, when you log in to anti-gravity for the first time, you're going to be prompted to log into your Google account here. And we want to hit this extension

here. And we want to hit this extension button right over here. And we can type in cloud code. Okay, this is how we actually download cloud code into anti-gravity. We'll need to make sure to

anti-gravity. We'll need to make sure to install it. And when we have it

install it. And when we have it installed, we can hit this button over here. And it's going to ask us to log

here. And it's going to ask us to log in. And this is where it costs money.

in. And this is where it costs money.

It's going to redirect us over here. If

we don't have a plan, we're I recommend starting with the pro plan and only going to the max plan if you absolutely need it. And once we have gone ahead and

need it. And once we have gone ahead and authenticated, it should take a couple seconds. We're going to have this window

seconds. We're going to have this window right over here. Okay. So, that's

essentially how we can get cloud code activated on our anti-gravity terminal.

We need to go ahead and literally close every single one of these tabs. We're

going to close this. We are going to close this. We're going to close this.

close this. We're going to close this.

And on the left hand side, I want to hit the documents over here. So, let's

quickly go over anti-gravity at a very basic level. Every single project

basic level. Every single project contains a root folder. Okay, this root folder is called NADN versus Claude and it contains all the subfiles and folders. Okay, so we need to go ahead

folders. Okay, so we need to go ahead and actually create that project by opening a new folder. I'm going to go into my documents here and just create a

brand new folder. And I'm going to call this clawed uh demo here. And it's going to be completely empty. That's totally

okay. We're going to open this file. And

now we have access to this parent folder right over here. When we're dealing with claude code, the most important thing to to do is planning. And I want to break

this down actually on my blackboard.

when you're dealing with um agentic workflows I think probably and don't quote me on this but I would say like 80% roughly of

the work you do comes in the form of something well planning okay so unfortunately a lot of people like skipping planning and going straight to action but if you're serious and you want to get the best results possible

you have to really focus a bit of time on on the planning okay it's kind of like back in the 16th century when you had all these um sailor and they're going from like Europe over to like

North America over here and they were charting a course trying to hit like I don't know somewhere in the States or Canada and they're on a path, right? The

problem is is they're trying to go straight but they have to have incredible navigation to hit it. Okay?

And if you veer off too far to one side, it doesn't seem like a massive distance, but as you keep going, you could hit like the Arctic Circle over here, and you could even go down to like South

America over here. And obviously that's really bad news. And so the reason I'm pointing this out is because if we compare it to cloud code, we want to plan properly so that we can get the

best results possible so that we can build out a workflow that is able to handle all of our year-end bookkeeping and accounting. And so how we do this is

and accounting. And so how we do this is we need to use something called a claw.md file. And this is just

claw.md file. And this is just essentially like an operations guide or like instructions manual on how we want it to get the results for us. This is

going to be down below in the description for you guys to copy. But

essentially, you'll head over to my free school community and you can download the cloud code uh claude.mmd file here.

Just copy everything and then inside of anti-gravity, we're going to hit new file. I'm going to call this claude.md

file. I'm going to call this claude.md

and paste it in. MD stands for markdown.

It's why it's like plain text with a bit of styling over here. You got blue over here. I'm just not sure if that's orange

here. I'm just not sure if that's orange or brown, but yeah, it's styling everything. Now, this is more or less

everything. Now, this is more or less the majority of the planning here. I'm

not going to go a mile deep into this, but I want to cover this at a surface level. First of all, our goal is to

level. First of all, our goal is to create uh working software by making small changes one step at a time. Before

we do anything, we actually have to define what it is we're building. And

we're going to define it in a project file called project specs. Name's

irrelevant. You can call it whatever you want. This describes to Claude what it

want. This describes to Claude what it is we're building. Just like if you hired an employee at a company, you're gonna want to tell that person what you're hiring them for, what their job is, what done looks like, what the

inputs are, outputs, all that kind of stuff. The more information we provide

stuff. The more information we provide to Claude or an employee, the better it's going to be at getting the job done. We can define any technology that

done. We can define any technology that we want to use here. So, if you are a programmer and maybe you want to use Python or JavaScript, which are two coding languages, you can define what it

is we're using. And then uh we just have some running the project steps here. The

file structure which is very important but we're going to be coming back to this. We're going to talk about other

this. We're going to talk about other things like how we want it to write the code, how we store secrets safely like passwords or API keys or all that kind of stuff and so on and so forth. Okay,

so that is essentially what this Claude MD file is is its instructions. Now with

that out of the way, let's go ahead and create our very first prompt inside Claude Code. I'm going to link to this

Claude Code. I'm going to link to this page in the free school community if you guys want to copy and paste everything.

The reason why I've standardized this in advance is because Claude code is probabilistic in nature. Let me explain what that means. If I send a message

like, "Hey, how are you today?" to

Claude 10 different times, I'll get 10 different responses. So, it will always

different responses. So, it will always slightly vary. So, what you get and what

slightly vary. So, what you get and what I get might be slightly different. So,

I'm trying to create prompts that will get us as close of results as possible.

Okay. So, we're going to go ahead. We're

going to copy this very first prompt over here. And then we're going to paste

over here. And then we're going to paste into Claude. So, I need to hit Claude

into Claude. So, I need to hit Claude over here. Okay. And we're going to

over here. Okay. And we're going to paste it right in here. Let me explain what is going on. First of all, we're going to tell it to read claude.md.

And we're going to ask it to build out the file structure that we created right over here. It's going to ask for our

over here. It's going to ask for our permission a bunch of times inside the cloud editor. Now, this can be somewhat

cloud editor. Now, this can be somewhat annoying if you're if you're not a programmer and you don't have any intention of actually going in and editing the code because you're going to

have to essentially approve it all the time. If you want to avoid this from

time. If you want to avoid this from happening, you can go ahead and change by clicking down here the mode to bypass permissions. And now Claude will just go

permissions. And now Claude will just go ahead and update everything as it sees fit. Really quickly here, just backing

fit. Really quickly here, just backing up, we are, you can think about us like the boss or the employer or the business owner. We're messaging our employee,

owner. We're messaging our employee, which is Cloud Code. And Cloud Code is going to have a conversation back and forth with us, telling us exactly what it's doing, every single step. This is

really important because with tools like NATM, there's a visual code editor which means that you can see every single step of the way and understand what's going on. But with Cloud Code, you don't have

on. But with Cloud Code, you don't have that. And so the next best thing is to

that. And so the next best thing is to understand every single step that Claude's actually taking because otherwise it just feels like a black box or magic where you don't really understand what's actually happening.

You send a prompt, you get a response back, but the in between is really unclear. with Claude, it's going to

unclear. with Claude, it's going to break that down and tell you exactly what's going on so you can see what's happening. And if you need to make any

happening. And if you need to make any edits, you can always message Claude down here and tell it what needs to happen or what needs to change. So, it's

read the the Cloud file. It's read the memory, which is just like memory of Claude. They create like context history

Claude. They create like context history for all of your conversations and it reads it, creates a to-do list down here on what needs to get done and then it just goes ahead and checks off every

single thing on the to-do list step by step. And it's gone ahead and created

step. And it's gone ahead and created that project file structure right over here. We're going to break this down

here. We're going to break this down very simply. Anytime we're dealing with

very simply. Anytime we're dealing with code that we're writing in this project, it's going to land in the SRC. So, for

example, if we're taking PDF documents like this and we're turning into text, that would be a code snippet, and it's going to add in SRC. If we're renaming uh PDF files, that's another code

snippet. If we're going to be organizing

snippet. If we're going to be organizing them into different files and folders, that's another code snippet, and so on and so forth. So, every action that we're asking to do is going to be a code snippet that belongs in the SRC. Next

thing is out. This is temporary files and folders. So, if we're asking for uh

and folders. So, if we're asking for uh cloud to create a CSV file, for example, it's going to land in there. Okay. Uh

and we'll get into that later.

Environmental the environmental variables folder. This is just password,

variables folder. This is just password, secret keys, all that kind of stuff.

Project specs.md is essentially the entire project. We're explaining what it

entire project. We're explaining what it is that we're looking for Claude to go ahead and build. And then

requirements.txt txt is just essentially all the packages that are required for claude to build the thing that we wanted to build. We should head over to project

to build. We should head over to project specs, read through this whole thing, and make sure we're happy with it. But

as a TLDDR, essentially, we just gave one prompt over here telling Claude what we wanted to build. It created this project specs, defining what this project does, what the inputs are,

outputs, how it works, the text stack, and all of the additional notes. I would

read through it. You can add in edge cases. you can add in what done looks

cases. you can add in what done looks like so you can define when it's good enough to be shipped into production all that kind of stuff. So it's good idea to practice this. But why this is important

practice this. But why this is important is because every time Claude is going to fire any code, it's always going to read these two files first to make sure that it's aligned with what it is we're

actually trying to build. Okay, so let's go down and take a look at the next steps here, which um everything looks

good. Now, what I want to do next is

good. Now, what I want to do next is head over to the second prompt here. So,

first of all, we're going to read the project file, and then we're going to build two actions in here. We're going

to take all of our invoices, and we're going to rename them properly in this format. Okay? The name of the company

format. Okay? The name of the company followed by the date. And then we're going to sort it into the month that that invoice actually came in. So, it's

going to organize our entire workflow.

We also want to dump this into a CSV file. Okay, so that we can um and let me

file. Okay, so that we can um and let me just pull out the CSV file so that we can see all of the information in a CSV format. Okay, so that's another benefit

format. Okay, so that's another benefit here. And we're going to take it a step

here. And we're going to take it a step further where we classify every invoice as either revenue or an expense. And if

it is an expense, we're going to break it down into the following categories.

So it's going to automate the bookkeeping. Is it a software expense, a

bookkeeping. Is it a software expense, a marketing expense, a professional services expense like lawyers, accountants, all that kind of stuff?

Like what actually is it? So, we're just going to define that very clearly. Now,

returning back to our project here, the first thing that I want to do is add in two PDFs. And I'm going to add all of

two PDFs. And I'm going to add all of the PDFs for you to go ahead and use so that you can follow along very clearly with me. But I'm just going to copy for

with me. But I'm just going to copy for this portion two PDFs in here. Now,

instead of using all 40 at the same time, we're using two because it's going to be substantially quicker. Now, I'm

going to copy in this next prompt here.

And on top of that, we need to add in an anthropic key, API key. The reason why is that right now, Cloud Code has access

to, you know, AI and and helping us do all this kind of stuff, but the actual act of, I believe, taking data out of a PDF is not something that's included in

this plan. So, we need the API key. So,

this plan. So, we need the API key. So,

we can head over to anthropic anthropic API key. And we'll click the first link here, create an account if we don't have one already. And then at the

bottom left, we'll click API keys.

Generate a new key. Let's call it something thoughtful and creative like that. And we'll copy it. Keep in mind,

that. And we'll copy it. Keep in mind, you do need billing enabled. And you'll

need to buy at a minimum $5 of credits in order to use this. Okay, cool. So,

I'm going to say here's my anthropic API key, and I'm just going to paste in here. Typically, you don't want to paste

here. Typically, you don't want to paste the API keys in this window here. You

want to add them into the MV file here to store them securely. I just ended up doing that to to be quick. And I'm going to delete that after this video anyways.

And again, just like before, we are the employ employer sending the message and the employee is writing back to us telling us exactly what it is that it's actually going ahead and doing. Cool.

So, everything looks good to go. It's

gone ahead made some updates over here and it's told us that everything looks good. Now, one thing is is that it's

good. Now, one thing is is that it's actually adding all of the code into one Python script, which is the main.python

script here. Again, you don't need to know anything in coding, but like all of these different functions, it's just building into one project file over here, and it's pretty long. We could ask

it to essentially, can you please break down every different task into a different Python script in the SRC? So,

we could go ahead and do that. And then

it would actually break everything up.

And then I'm going to say, can you please run the scripts to process the

two PDF files I have in the PDFs folder.

Okay, so we're getting it to do two things. Breaking it down by the script,

things. Breaking it down by the script, and then we're going to actually rename, reorganize, and also dump these invoices into a CSV

format. And one thing that I find

format. And one thing that I find hilarious is the words that um Cloud Code is just using down here. It puts a massive smile on my face. Like it's

saying concocting, manifesting, the other word I couldn't even pronounce, philosophizing, all these kind of things that's just absolutely wild and crazy

and silly, musing. It just never ends.

But anyways, we can see that Cloud Code has actually gone through the process of taking all of these different tasks and putting them into their respective um

their respective files. So, we have for example a uh code snippet to export data to a CSV file. We have a code snippet to extract text from a PDF. We have a code

snippet that organizes the PDF. We have

a code snippet that could flag certain high invoice amounts and we actually had it in this particular prompt here. But

we also said flag any duplicate invoices or any invoices that seem unusually high. So if we get like a software build

high. So if we get like a software build that's like astronomically high, it'll let us know that. Okay. And now it's going to go ahead and run this. And one

other thing here, and you can see it coming in, is that it's fixing all of the code. So Claude is actually making a

the code. So Claude is actually making a whole bunch of mistakes and it's actually updating everything. Now the

reason why that's really important is because if you're using a tool like any you were on the hook for debugging and as you probably know if you use a tool like this the last 20% is 80% of your

time. So building out this workflow, you

time. So building out this workflow, you can pull all these nodes really quickly, but actually debugging, making it making sure everything works is the largest amount of your time and you have to be

technical enough and knowledgeable enough to fix it. But with cloud code, it just goes ahead and it fixes everything for you. Okay, so it looks like everything's done and we have our

um files now in the out folder.

Remember, this is every time we output data, it's going to be here and it's broken down by the month. We can see that both of these are from November 2025. And so it's broken them down. And

2025. And so it's broken them down. And

it's also it would have gone ahead and relabeled them. Now I actually went

relabeled them. Now I actually went ahead and already had these all organized properly. But if this was like

organized properly. But if this was like scrambled and disorganized, then it would of course go ahead and reabel that. Okay. Awesome. So we've gone ahead

that. Okay. Awesome. So we've gone ahead done a whole bunch of things here. First

of all, we've gone ahead and we've reorganized them into their respective months. Renamed them. We've also gone

months. Renamed them. We've also gone ahead and dumped that into a CSV file right over here for us to go ahead and export. And it's also gone ahead and

export. And it's also gone ahead and told us anything that was flagged. And

it says both PDFs dated whatever for our the same invoice. So this is actually a duplicate invoice that I've gone ahead and put through and we've sorted it.

Okay. So I could go ahead and ask Claude Code to essentially remove that duplicate as well. So that's another great thing that it can do. and it's

told us that there's no high amount invoices. Both of these look normal.

invoices. Both of these look normal.

Okay, this is great. But so far everything's on our computer. Let's go

ahead and let's take the next step, which is going to be uploading all this data into Air Table. Again, we're going to just use two invoice examples here and then we're going to add the next 40 on top of this. So, I'm going to add in

the next prompt. Essentially, this is just saying take all the CSV data or everything we've done so far, including all these fields, and upload this into Air Table. Now, in order to upload

Air Table. Now, in order to upload things into Air Table, we do need something called a token. It's like a password or a secret key or whatever to actually authenticate with Air Table that we are who we say we are and we have this c certain permissions to do

the things that we want to do. Now, you

again could go into um you could go into cloud code and ask it how to walk you through or like how to set up the uh token, but I'm just going to go do this really quickly because I've done this a

million times. We're going to head over

million times. We're going to head over to airable.com/create/tokens.

to airable.com/create/tokens.

We're going to create a token here.

Going to give it a really fancy name.

And we are going to add in four different scopes here. I'm going to first add them. Then I'm going to explain what's going on here. So inside

air table, okay, we're giving it access to read and write both the data and the schema. So if we open up air table, this

schema. So if we open up air table, this is like a table here. We're giving it permission to read everything, see what's actually there, and then add or update. Similarly, we are giving it

update. Similarly, we are giving it access to read the schemas over here and add or update everything. Okay, so that is essentially it. And we also want to go ahead and make sure that we give it

access to all of our resources instead of just one table or whatever you you of course you can choose whatever permissions you want. I'm just giving it access to everything. We can copy that

and I'm going to say my token is this.

I'm going to paste it off and let Claude code do all of the heavy lifting for me.

First of all, it's going to go through.

It's going to find this table. It's

going to analyze all the columns. If

they don't exist, it's going to create them. And then it's going to add all of

them. And then it's going to add all of that data in. I'm going to first start off by deleting this just so we can start nice and fresh here with new data that's coming in from this particular

build. Schlleing. What a what a word to

build. Schlleing. What a what a word to use. I uh don't think I've ever used the

use. I uh don't think I've ever used the word schluffing [laughter] until I said it live on camera right here because Cloud Code put it there.

Perusing. What another word choice, man.

Have you ever used the word perusing in your life? Maybe that is something that

your life? Maybe that is something that I should be using more. Sweet. And so

everything is looking good. It has gone ahead and it's just completing everything. Let's see if the data is

everything. Let's see if the data is coming in here, which it looks like it is. Okay, everything looks good so far,

is. Okay, everything looks good so far, which means that we are officially good to go. We have added all this data into

to go. We have added all this data into Air Table. Now, I'm actually just going

Air Table. Now, I'm actually just going to go ahead and this time I'm going to add in the remaining 40 or so invoices into my um window here. I'm going to

pause the video. I'm going to process all of these and return back once it's done. All right, so it looks like it's

done. All right, so it looks like it's done. And uh just wrapping up here, but

done. And uh just wrapping up here, but I'm just going to go back up to the prompt that I typed in here. I

essentially said, "Do everything for me." I want you to take all of these

me." I want you to take all of these PDFs, all 42 of them or whatever. And

then I want you to essentially export them into the right folder depending on the month. And I only have invoices from

the month. And I only have invoices from the from August to December. And I want you to rename them, but this time rename them in a way where the date comes first followed by the service provider. So you

can see that it actually flipped everything, sorted it properly, then it added everything into this CSV file and it also added everything into this

particular air table um data database as well. Now another cool thing worth

well. Now another cool thing worth mentioning here is that not only did it do all of that, but we can see that it edited a whole bunch of stuff and it

encountered a ton of mistakes along the way. And the cool thing is is that this

way. And the cool thing is is that this whole thing was one prompt. like I

didn't have to reprompt anything or give it additional instructions. Our junior

or senior developer, I guess in this case, literally went through the whole thing and got the entire job done for me with one single prompt. Now, inside NAD, if you were to do this and encounter

those 10 or 20 mistakes that cloud code encountered, you'd be on the hook for doing all of those 20 mistakes. And we

both know that that would take a substantial amount of time, whereas with cloud code, it was just a couple of minutes. Okay, cool. Now, there's just

minutes. Okay, cool. Now, there's just one mistake here that I noticed, and that is that it's labeling a category for every single invoice when I only

want it to be on um expense invoices.

And I also don't want everything to be an expense because well, there's some revenue sources in here as well. So, I'm

going to go back to the drawing board and say, I noticed that you labeled all invoices as expenses

and put a category on everything.

Categories should only be on expenses.

Can you make sure to classify

each invoice uh PDF properly as either uh revenue or expense? Okay, cool. So

again, just very basic prompt there and it's going to go to work for us updating hopefully the entire air table without us having to do anything else. Cool. And

so it is just wrapping up here and it looks like it's accomplished everything again just in one prompt. So essentially

what it was able to do if we look at air table now is it's gone through okay and it's broken things down by expenses and revenue okay and it's removed the

categories for revenue and it's only kept it for expenses. Now, we actually have two type columns here. At the end of this, I will just go ahead and delete the second one here and then just only

use this moving forward. Okay, so that's awesome. Now, on to the last prompt that

awesome. Now, on to the last prompt that we're going to go over in this video is going to be actually graphing this. And

again, this is something that NAD is not really that capable of doing for us. So,

specifically, what we're going to be doing here is we're asking Cloud Code to read the CSV file and create two separate things. Number one, it's going

separate things. Number one, it's going to create a dashboard for us. Okay, very

similarly to this dashboard right over here. Then number two is it's going to

here. Then number two is it's going to create a summary file where we can see all the information that we care about.

So for example, total revenue, expenses, net profit, tax collected so that we know how much tax we owe at the end of the year, the top 10 clients that we have so that maybe we double down on servicing them, making sure they're

happy so that we can continue to make as much money as we are. And we can break it down by revenue [clears throat] by month or in this case we could actually do a profit loss statement. So we could

see revenue and expenses every single month broken down. And then we're going to actually show this in an HTML dashboard so that we can see um graphs

and charts so that it makes sense and we can represent the data. Well, cool. So,

we're just wrapping up on this prompt here. And I'm going to uh z I'm just

here. And I'm going to uh z I'm just going to go up in the actual channel here. I did run out of memory in this

here. I did run out of memory in this conversation. So, [clears throat] I just

conversation. So, [clears throat] I just hit this command which is /compact.

Essentially, all we're doing is that we are condensing this whole chat window, which obviously is a lot of information.

We're just condensing it down into a smaller message here moving forward so that Claude hasn't having to remember the entire context of all past messages.

Okay, so we've gone ahead, we've compressed the chat and it's gone ahead and ran everything for us. The first

thing is is it's created this summary file right over here. So we can see for example a profit and loss statement, total revenue, expenses, net profit, taxes collected, top 10 clients by

revenue, as well as the revenue of the month. Now, pretty basic stuff, but of

month. Now, pretty basic stuff, but of course you could add way more on to this if you spent a bit more than a couple minutes building this out. And the next

thing here is um we can go ahead and take a look at our um at our dashboard here. So let's go ahead and see this.

here. So let's go ahead and see this.

Okay.

And here we go. We can see that it's created a dashboard. This is on our computer right now. Technically, we can see the revenue, total expenses, net profit, tax collected. And we can see

revenue and expenses by the month, which I think is really cool. We can even see expenses broken down by category, and top 10 clients. The cool thing is is now we can ask clawed questions like for

example, is there any expenses that I could potentially cut from what I'm spending money on? And hopefully that could save you money. And again, so many additional things that we could do. We

could ask questions about certain months or whatever the case may be. We can

connect this into QuickBooks, remove duplicates, all of that kind of stuff.

And just the last thing that I put in this guide here was that you can type this last prompt into Claude. And what

this does is it actually will water watermark all of your invoices with this processed stamp on here, which I think is really cool. Now beyond just that another big difference is is that

technically this whole workflow is on our computer whereas when we look at NADN this is all on the cloud but you can also go ahead and upload this entire workflow into the cloud so that you can

run it from anywhere anytime by anyone using apps like for example modal.com or for example trigger.dev dev or for

example replet for example or render.

These are like four different applications. And again, there's way

applications. And again, there's way more than just these four, but the point is is that you can actually upload this in the cloud and have it running all the time. So that's it for this video, guys.

time. So that's it for this video, guys.

The blueprints for this are going to be down below. Again, in the school

down below. Again, in the school community right over here, we have this GitHub project. All you have to do is go

GitHub project. All you have to do is go into it, hit the code button here, copy this, and enter in clone this project and then paste that link and it will

literally give you everything here right away. If you guys want to learn more

away. If you guys want to learn more about AI and automation, I highly recommend taking a look at my school community where there are three major outcomes. Of course, you can learn tools

outcomes. Of course, you can learn tools like cloud code, anti-gravity, and all that kind of stuff so that you can save time and earn money. Number two for transformations is for those of you that are looking to create an AI automation

agency or become a freelancer selling AI automation services to business owners.

I'll show you how you can find, close, and fulfill your first deal in one month or less and then repeat the process for future clients. And then lastly, if you

future clients. And then lastly, if you guys have an existing business, I'll show you how you can automate up to 80% of what you're doing manually by using the blueprints that I have in this community. You can just copy, paste, and

community. You can just copy, paste, and deploy them. And these are the exact

deploy them. And these are the exact blueprints that allowed me to scale to seven figures. And if you guys do not

seven figures. And if you guys do not want to have to build these solutions out yourself, then you can always go ahead and take a look at my um AI automation agency where we can build it out on your behalf. Thanks guys for

watching and I'll see you in the next