HomeVideos

Grok Bot Just Got 70% Cheaper... (I Switched)

Now Playing

Grok Bot Just Got 70% Cheaper... (I Switched)

Transcript

765 segments

0:00

This might be controversial, but I was

0:02

completely wrong about Grockbot. I've

0:03

been using it every single day since it

0:05

released about 12 days ago, and my view

0:07

on it has completely flipped. On top of

0:10

that, it just got 70% cheaper to use

0:12

today, and you could try it out for

0:14

free. So, in this video, I'm going to go

0:15

over real life ways that I'm using this,

0:18

what I like about it compared to other

0:19

AI agents like Claude Co-work, and I'm

0:21

even going to give you tactical things

0:23

that can give it superpowers that most

0:25

people don't know about. So, by the end

0:26

of this video, you'll understand if

0:28

Grockbot is right for you. Let's dive

0:30

in. Now, first of all, we have some

0:31

great news. If you've been wanting to

0:32

try out Grockbot, but you didn't because

0:34

it was too expensive. They just made it

0:36

accessible to some of the cheaper plans.

0:38

So, let me go over how you could use

0:40

this right now and how much you need to

0:41

pay in order to do so. Now, here's the

0:43

important part. I do want to point out

0:44

that they're offering a free trial with

0:47

limited usage for all other users. So,

0:49

this probably won't get you too far, but

0:51

it will at least give you the ability to

0:53

test this out yourself and see what it's

0:55

capable of. If we take a look at pricing

0:57

right here, we could be on the cursor

0:59

pro plan for $60 per month. And with

1:02

this, we get access to Grockbot. Before

1:05

today, we would have to pay $200 per

1:07

month, which is extremely steep for

1:09

people to go and try this out. So, I

1:10

really do think this price change is

1:12

going to make it much more accessible

1:13

for everyday people to begin using this.

1:16

Now, if you still don't want to fork out

1:18

$60 per month, make sure to stay to the

1:20

end of this video because I'm going to

1:21

go over everything that Grockbot is

1:23

capable of. So, you don't have to guess

1:25

on whether or not this is going to be

1:26

worth it for you or not. Now, with

1:27

pricing out of the way, let's dive into

1:29

how exactly to use this, where we can

1:31

use it, what the platform actually looks

1:33

like, and then we're going to dive into

1:34

the agents and use cases that I've

1:36

personally been using every single day

1:38

for the last 10 to [music] 12 days. So,

1:40

first of all, you can go ahead and

1:41

download this. I'm going to add a link

1:42

in the description. You could download

1:44

this for your Mac operating system as

1:46

well as you could download this on your

1:48

phone, which is my personal favorite way

1:50

to interact with this. Now, once you

1:52

download this onto your desktop and you

1:53

open it up, this is what you're going to

1:55

see. Well, I guess you won't see all of

1:57

this because I have ton of different

1:59

agents I've configured. Let me break

2:00

down exactly how this works. So, one

2:02

thing that is fundamentally different

2:04

between this and something like Claude

2:06

Code, Claude Co-work, Codeex, and all

2:08

these different AI agents that we've

2:09

used in the past is that we have

2:11

multiple different agents that are

2:13

basically just chat threads that we can

2:15

interact with. As you can see, I have a

2:17

health agent up top. I have my main

2:19

agent. I have a brand partnerships

2:21

agent, contentbot, short form, inbox

2:24

triage, investment agent. I basically

2:25

have probably at this point like 12

2:27

different agents that I'm running every

2:29

single day. And if I click on one of

2:31

these agents, you could see all of the

2:33

previous conversations that we've had

2:34

with it, which I personally like. It

2:36

makes it very organized. So when I want

2:38

to go and do anything email related, all

2:40

I have to do is go to my email agent and

2:42

I could see everything from previous

2:44

conversations. Now, one thing I do want

2:46

to point out is as at the top you can

2:49

see I have all of my demo agents. These

2:51

are the same exact agents that I've

2:53

configured that I run for myself

2:55

personally, but I told these agents to

2:57

use different names for clients and

2:59

different price numbers. So, all of this

3:01

is real information, but it's not the

3:04

actual clients in the price because I

3:05

didn't want to show this for

3:06

confidential reasons. And to be honest,

3:08

it's kind of hard to show how I'm using

3:10

Grockbot when it has all this

3:12

confidential information. So, think of

3:14

these demos as the exact same process

3:16

that I use, but not with my real

3:18

information. One thing I want to point

3:20

out is not only can we have specific

3:22

conversations with each of these

3:23

different Grock bots, but they can speak

3:26

with each other. So, if I pull up my

3:28

main agent right here, you can see it's

3:30

messaging with two other separate

3:32

agents, which is pretty cool. I didn't

3:34

even have to go and prompt it to do so.

3:35

It just knows whenever there is

3:37

something brand partnerships related, it

3:39

will go and text that agent. And the

3:40

same thing for my email agent. If I ever

3:43

want to see what they're saying back and

3:44

forth with each other, I could just

3:46

click on this. And I'm going to click on

3:47

email agent. And I mean, take a look at

3:49

this. Our main agent gave a brain dump

3:51

on basically everything that I want

3:53

accomplished. From there, my email agent

3:55

understands this goal and is able to go

3:57

and pull all this relevant information

3:59

for me and then feed it back to the main

4:01

agent. Then inside of my main

4:03

conversation here with the main agent,

4:06

it breaks all this down as you could see

4:08

right here. So, this is a fundamentally

4:10

different way that we interact with AI

4:13

that I'm actually really starting to

4:15

love. Now, next up, what I want to talk

4:16

about is plugins because this is how I'm

4:19

able to access things like ClickUp or

4:21

Gmail or Granola and look at all my

4:24

meeting notes and all those different

4:25

things. And this is where it could get

4:27

really powerful when we connect it to

4:28

the different applications that we're

4:29

using on a day-to-day basis. So, if I

4:32

come over here on the left-hand side,

4:34

you could see this little button that

4:35

says plugins. So, I'm going to click on

4:36

this. I have a couple that I've already

4:38

preconfigured, but you could go through

4:40

and search through this entire list and

4:42

add some of the different apps you use.

4:43

I think right now there's maybe like a

4:45

couple hundred pre-builtin connectors

4:48

that you could go ahead and sort

4:49

through. Now, the best part about the

4:51

plugins is we don't even necessarily

4:52

need to go and configure these

4:54

ourselves. We could just come into our

4:55

chat and simply just say, "Hey, I want

4:57

you to connect to my Google calendar.

4:58

Can you do it for me?" And Grockbot is

5:00

really good at doing the entire process

5:02

end to end. All we have to do is

5:04

authorize the account, which it'll

5:06

populate with this little card here for

5:08

us to click on, and it'll just do it for

5:10

us automatically. Now, let's talk about

5:11

use case number one that I've been

5:13

using. Rockbot 4 and that is to help me

5:15

with my email inbox and this has been a

5:18

gamecher for me personally. Basically, I

5:20

configured my main agent to add a

5:22

routine which is basically a scheduled

5:24

task that could run at any given time.

5:26

And so every morning at 8 a.m. it

5:28

automatically goes and scrapes not only

5:30

my personal email inbox, but also my

5:32

sponsorship email as well since we can

5:34

connect multiple different Gmail

5:36

accounts to our agents. And then it

5:38

basically just gives me the three things

5:39

that I need to focus on today. if

5:41

there's a specific contract that needs

5:42

to get signed or if there's an invoice

5:44

that needs to be generated, then the

5:46

agent will update me with that. But a

5:48

key thing that I want to note here is

5:49

I'm interacting with the main agent, but

5:52

this is communicating on my behalf with

5:54

the email agent right here. So, if I

5:56

want to open this up and see kind of

5:58

what they're talking about, our main

6:00

agent just says, "Hey, let's run our

6:01

scheduled task for the morning inbox

6:04

check." And then the email agent goes

6:05

and pulls all this information. And then

6:07

my main one comes back with all the

6:09

three things that I need to focus on and

6:11

do today. Now, in order to set up a

6:13

routine like this, it's very simple.

6:14

Either you could come over to the top

6:16

right, like you could see right here

6:17

with this little computer, and then we

6:19

have routines that we could preconfigure

6:21

right here. All you have to do is click

6:22

add. You could give it a name,

6:24

instruction, and you could click the

6:25

trigger, like basically when you want

6:27

this to run, or you could just come into

6:29

the chat conversation, say, "Hey, I want

6:30

you to run this task at this given time

6:33

and this given day." and it will set

6:35

that up for you automatically. Now, one

6:37

thing I've been really loving about

6:38

using Grockbot is the fact that we can

6:40

add multiple accounts to each of the

6:42

different apps that we connect. So, if

6:44

you do want to add multiple email

6:45

inboxes, I'm going to show you exactly

6:47

how to do that right now, cuz I get tons

6:49

of people asking how you could do this

6:50

inside of Claude, and it's simply just

6:52

not possible. This solves a major issue

6:54

for people that are juggling multiple

6:56

different email inboxes at once. So, if

6:58

I click on the Gmail plugin, I have two

7:00

different accounts. I have my default

7:02

account and my sponsorship email account

7:04

added. And all I have to do is add

7:05

another account here. Give it a name.

7:06

I'm just going to call this test or

7:08

something. Click authorize. And then

7:09

what it's going to do is it's going to

7:10

pull up your Gmail. And all you have to

7:12

do is just authorize that account. And

7:13

it takes one click to do so. Then if I

7:15

come into my email agent, I can just

7:17

say, "Hey, can you give me the last

7:18

three emails from both of my Gmail

7:20

accounts, please?" And just like that,

7:22

we get this response. It pulled from my

7:24

personal account as well as my other

7:25

inbox. And again, this has been

7:27

incredibly helpful for helping me manage

7:29

multiple inboxes at once, which we

7:31

weren't able to do with other AI tools

7:32

like Claude, Hermes Agent, OpenClaw,

7:35

etc. Now, this next use case is one

7:37

that's been really helpful in my life,

7:39

both personally, but on a business side,

7:41

and this is my meeting agent. So, let me

7:43

quickly break down the flow of this

7:45

right now before showing you a demo. So,

7:47

at the end of every single day, I have a

7:49

routine that fires at 5:00 p.m. It will

7:51

then automatically pull all of my

7:53

Granola meeting notes, which if you

7:55

don't know what Granola is, it's

7:56

basically an AI notetaker where it

7:58

automatically joins all of my meetings,

8:00

gives all the action items from those

8:02

meetings, and then inside of the meeting

8:04

agent, it will automatically add those

8:06

action items into ClickUp, and assign

8:08

the due dates as well as the team member

8:10

that needs to complete these tasks for

8:12

me. So, for example, I have this school

8:15

community board inside of ClickUp, and

8:17

literally all I did is have a

8:19

conversation with Granola and give a

8:21

brain dump on all the things that I need

8:23

to do inside of my school community. And

8:25

then it automatically added all of this

8:26

to my to-do list as well as it dragged

8:28

the specific tasks into the in progress

8:31

tab as well. Now, first of all, if you

8:33

want to sign up for Granola, there is a

8:34

link in the description. You could

8:36

download this for your desktop as well

8:38

as on your phone. And then once you have

8:40

this downloaded, you're going to see

8:42

this interface here. So I personally

8:43

have all these different meetings coming

8:45

up which Granola is automatically going

8:47

to join without me needing to like tell

8:49

it to go and join these meetings. And

8:50

then here are all the different notes

8:52

that I have. These are a mix of

8:53

different meetings that I had with team

8:55

members for my business, but also just

8:57

like notes that I had. So like yesterday

8:59

I went on a run and I basically broke

9:01

down in a voice note for about 5 minutes

9:03

everything that I need to do and

9:05

prioritize over the next week before I

9:07

go on this trip. Now, if I go and click

9:09

on one of these notes, you would see

9:10

that it automatically took the

9:12

transcript and then summarized it and

9:14

then added these bullet points right

9:15

here. That makes it really easy for not

9:17

only me to understand what this meeting

9:18

was about, but also for my EI agents to

9:20

understand as well. Now, the best part

9:22

about the meeting agent and granola

9:24

combined together is I have this

9:27

scheduled task where every single day at

9:29

5:00 p.m. it's automatically going to go

9:31

and pull all the meetings that I had

9:33

inside of Granola. And then what it's

9:35

going to do is it's actually going to

9:36

add these to a specific board inside of

9:40

ClickUp in order for me to track what's

9:41

going on inside of my business. So, as

9:43

you can see, it automatically added

9:44

these things to this specific board

9:46

inside of ClickUp. And one thing I want

9:48

to note is not only does it add it to

9:49

the board, but it picks which board to

9:51

add it to. Cuz for example, I have Loom

9:54

campaigns, which is for my agency, I

9:55

have my invoices and my payments, and I

9:57

even have a Grockbot task tracker here.

10:00

So, there's multiple different boards

10:02

that it can add to, but it knows

10:03

specifically which one to go and add

10:05

these tasks to. And notice that I said

10:07

that this is in progress confirming

10:08

YouTube shorts metadata in workflow

10:10

inside of Metrocool. So, whenever me and

10:12

my business partner are on a meeting or

10:14

something, it'll automatically like move

10:16

these to whatever process this is in for

10:18

our business. Now, let me show you how I

10:20

could use Granola on my phone and

10:22

basically just give it a brain dump and

10:23

then it will automatically use my

10:24

meeting agent to then route this to

10:26

ClickUp. Right here, I'm just going to

10:27

come and click add a new note. Hey, so

10:29

I'm working on a couple of things right

10:31

now um that are kind of top of mind. So,

10:33

first of all, I'm finishing up this

10:34

Grock Bob video. So, I need to confirm

10:37

the packaging. I need to then generate

10:38

the thumbnail for this video. On top of

10:40

that, I need to confirm that my flight

10:42

is rebooked to Lisbon cuz that flight is

10:44

in a couple of days. Also, can you add a

10:47

note to make sure that I don't forget my

10:48

passport before going to the airport?

10:50

So, as you can see, I just simply

10:52

recorded what I was talking about. You

10:53

could let this thing run for like 30

10:55

minutes and give it a brain dump if you

10:56

want. And then what it's going to do is

10:58

it's going to generate a TLDDR breakdown

11:00

of what it is that we talked about. So

11:02

just like that, there's those bullet

11:03

points right here. I could also access

11:04

this on my computer as well since it

11:06

syncs up with my desktop. Now, if I come

11:08

into my meeting agent inside of Grockbot

11:10

and I click on this computer up here and

11:12

I click on the routine, I could just

11:14

test run this right now. So what it's

11:15

going to do is it's going to scrape

11:17

Granola via the connector that I gave

11:18

it. And then it's going to give me the

11:20

action items and add these to ClickUp

11:22

for me. Right? So like 30 seconds later,

11:24

it then scraped this new note that we

11:26

just took. It found my YouTube video

11:27

board and added this right here that I

11:29

need to confirm Grockbot video packaging

11:31

and thumbnail generation as well as

11:33

inside of my Loom board. It said pack

11:35

passport before leaving for Lisbon. Now,

11:37

let's say you scroll through this entire

11:39

plug-in marketplace and you don't see an

11:41

app that you want to use that you want

11:43

to integrate your agent with. Well, you

11:44

can use something called the Zap Your

11:46

MCP, which allows you to connect to

11:48

9,000 plus different apps. And it's

11:49

really easy to set up. So that way you

11:51

can work across more applications than

11:53

just the ones that Grockbot allows you

11:55

to. For example, I have this MCP server

11:58

configured right here which has about

12:00

like 25 different applications that I

12:02

can use. For example, there's some apps

12:04

in here that I cannot connect to

12:06

directly inside of Grock. So I cannot

12:08

connect to school which is where I have

12:09

my entire community as well as Beehive

12:12

is another big one which is where I run

12:14

my email newsletter. So, if I want to be

12:16

able to perform actions inside of

12:18

Beehive with my agents that I configure

12:20

inside of Grockbot, I actually have to

12:22

be using Zap Your MCP since it's not

12:24

accessible in this list right here. Now,

12:26

this is very simple to set up. First of

12:28

all, you just need to have a Zapier

12:30

account. You could sign up with a link

12:31

in the description and then you're going

12:32

to come into your plug-in marketplace

12:34

and then simply just look up Zapier and

12:36

then click on this add button. It's

12:38

going to ask you to authorize your

12:39

account from inside of Zapier. And then

12:41

it will show that it's added right here.

12:43

And in order to test this out, let me

12:44

just come into my inbox agent and say,

12:46

"Hey, I want you to use the Zap Your MCP

12:48

to pull different stats from my Beehive

12:51

campaigns so I can understand how many

12:53

people I have on my newsletter." And

12:55

here we go. It went and pulled from

12:56

Beehive using the Zap Your MCP. It shows

12:58

a newsletter size. It shows my open rate

13:01

as well as the average click rate. And

13:03

the cool part is I could even have it go

13:04

and perform actions inside of Beehive

13:06

for me since it's now connected. Now,

13:08

Grockbot or any AI agent for that matter

13:10

is only as good as the data that we feed

13:12

it. So, if we're trying to scrape data

13:14

from the internet, we want it to be

13:15

efficient, fast, and verified from

13:17

sources. Well, I'm going to show you a

13:19

simple plugin that you can add to

13:21

Grockbot. So, that way, you never have

13:23

to worry about whether or not you're

13:24

getting the right information inside of

13:26

Grockbot. All right, I want you to think

13:28

about it like this. A normal search with

13:30

a tool like Grockbot is basically like a

13:32

librarian handing you a cart full of

13:34

books. You personally still have to read

13:37

all these books yourself. Well, if we

13:38

use a tool like Parallel to read the

13:41

books, it will then just hand you the

13:43

three or four different lines that

13:45

actually matter instead of us needing to

13:46

go and source through all this

13:48

information ourselves. So, this does

13:50

three different things for Grockbot.

13:52

One, it's faster. We don't need to have

13:54

Grockbot go and click through these

13:55

different pages. Two, it uses fewer

13:57

tokens, so our Grock plan lasts us

14:00

longer. And three, it's just more

14:02

accurate since our agent is working from

14:04

the actual source material. It just

14:06

makes all of our answers inside of

14:07

Grockbot more accurate. Not only that,

14:09

but if you plug in parallel to your

14:11

Grockbot, you can just have confidence

14:13

that it's going to be able to go and

14:14

scrape whatever it is that you want.

14:16

Now, another way that parallel can be

14:18

extremely useful when you tie it with

14:19

Grockbot is having it go and monitor a

14:22

specific web page. So, for example,

14:24

let's say you have a competitor and you

14:26

want to track their pricing. You could

14:27

have it go monitor that website and when

14:29

there is a new change to the pricing

14:31

could automatically run a deep research

14:33

task inside of parallel and then you

14:35

could have Grockbot text you the final

14:37

information. And that's just one use

14:39

case. There's plenty of other ones you

14:40

could do in order to monitor specific

14:42

web pages on the internet. Now in order

14:43

to give you an example of how this deep

14:45

research with parallel works inside of

14:47

Grockbot I want to show you the output

14:49

that it gets when it does a deep

14:50

research task for us. So for example it

14:52

shows right here give a history on open

14:55

web. very simple prompt and what it did

14:57

is it gave this research document back

15:00

and this is what the human can view.

15:01

This is like all the data that it

15:03

extracts and shows it in this interface

15:05

for a human. But what we want to focus

15:07

on is if I click on machine here, this

15:09

is what our agent is then fed and this

15:11

is a more efficient way for our agent to

15:14

store data in order for us to interact

15:16

with it. Now let me show this in action.

15:18

So, what I did is I basically came to my

15:20

agent and said, "I want you to use

15:21

Parallel in order to monitor the price

15:23

of Bitcoin. Once it hits $80,000, I want

15:25

you to message me directly here inside

15:27

of this agent." So, what it did is it

15:29

went ahead and set this up inside of

15:31

Parallel. I didn't even have to see any

15:33

of the work that it was doing. But, if I

15:34

open up my Parallel account and I click

15:36

on monitor, you can now see that we have

15:39

this query right here. So, it's going to

15:40

use this monitor task in order to track

15:43

the price of Bitcoin on coin gecko or

15:45

coinarketcap.com. This is running every

15:47

single hour and when it hits that price

15:50

of $80,000 or if it ever does, it's then

15:52

going to text me through my agent. And

15:54

this is just one simple use case that

15:56

you could use when using Parallel, but

15:58

it's really powerful when you combine it

15:59

with Grockbot or any of the agents that

16:01

you're using. And it's super simple to

16:02

set up. So, let me show you how to set

16:04

it up right now. First of all, you're

16:05

going to want to sign up with an

16:06

account. If you don't have an account,

16:08

there is a link in the description to

16:09

sign up. And then all we need to do is

16:11

come into Grockbot and either you could

16:13

just chat with the agent and say, "Hey,

16:15

I want you to connect Parallel to our

16:18

account or we could come over to plugin

16:20

here and then simply search up

16:21

Parallel." I already have mine added,

16:23

but let me show you how to install it if

16:25

you don't already have it installed.

16:26

Now, it's as simple as clicking this add

16:28

button here. It's then going to have you

16:30

authorize your account in the account

16:31

that you just set up. And then just like

16:33

that, we are all set. Now, this next use

16:35

case is one specific to me and the type

16:38

of content that I make. But the main

16:39

point of this is that we could use

16:41

Grockbot in order to scrape X for

16:43

content ideas or just trending things

16:45

going on in the AI space. So, we could

16:47

almost use Grockbot to filter out

16:50

different things that are going on,

16:52

whether that's inside of our industry

16:53

inside of AI, your specific industry,

16:55

etc. So for example, I just gave it this

16:57

very basic prompt inside of my content

16:59

bot and said scrape X over the last two

17:01

days and find trending things in the AI

17:03

space that would be good for content

17:04

ideas. Then since it directly connects

17:07

to Twitter or they call it X now it

17:09

automatically will go and find all the

17:12

things trending. So that way I don't

17:13

need to spend my time scrolling on X

17:15

wasting time because it could just suck

17:17

me in and eats up all my productivity.

17:20

Now, in order to set this up, you're

17:21

going to come over to plugins right

17:23

here, and then you could simply just

17:24

search for X. And it's as simple as just

17:27

clicking authorize, and it will set this

17:28

up for you. Or you could just come to

17:30

chat and say, "Hey, I want you to add

17:31

X." So, you could pull all the posts

17:33

from there. Now, this next use case is

17:35

one that I'm personally going to be

17:36

using quite a bit, and I've really been

17:38

loving playing around with this over the

17:40

last day. And that is my invoice agent.

17:42

And not only am I just going to use this

17:44

invoice agent, but this is going to be

17:45

paired with my inbox agent that I just

17:47

showed you earlier on in the video. Now,

17:49

let me quickly break down exactly how

17:51

this is going to work. So, I have my

17:53

invoicing agent, and what I'm going to

17:55

do is I'm going to set an autot trigger,

17:57

which is one of the coolest parts of

17:58

Grockbot, which I'm going to explain in

18:00

a little bit. And what this means is

18:02

whenever an invoice email arrives, it's

18:04

automatically going to flag this

18:06

automation where it's going to then

18:08

draft a invoice in my style, send the

18:11

email to this client, and then on top of

18:13

that, it's going to track this in

18:15

ClickUp, so I could see exactly what

18:17

stage my payments are in inside of

18:19

ClickUp, as well as my entire team can

18:21

view that as well. Now, let me show you

18:23

exactly how to set this up, cuz it's

18:24

really easy to create a new bot or an

18:27

agent inside of Grobbox. So, I'm just

18:29

going to click on create new. And all

18:30

I'm going to do is I'm going to come

18:31

down to messagebot and just give it a

18:33

brain dump. All right. So, here's the

18:34

prompt that I'm going to give it. I

18:36

basically said create an invoice agent

18:37

that does these tasks. Works with the

18:40

inbox triage agent to scrape my emails.

18:42

Keeps track of all invoices and owed

18:44

payments. Generates invoices for me and

18:46

a specific style. And then tracks the

18:48

status of all invoices inside of my

18:51

ClickUp board called invoices and

18:53

payments. Then next, what I'm going to

18:55

do is I'm just going to upload a PDF

18:57

example of what I want this invoice to

18:59

look like. So that way it always

19:00

recreates it in this style with all of

19:02

my payment information. So I never need

19:04

to basically generate an invoice myself

19:06

again. So here is an example right here

19:08

of an invoice that I wanted to recreate.

19:10

I really like this style. And now when I

19:12

upload this as a reference, it's always

19:14

going to look exactly like this. So what

19:16

I'm going to do now is simply just send

19:18

this off to my agent. And now, as you

19:19

can see, it's going and crafting this

19:22

invoice style as well as just setting up

19:24

this entire agent. One thing that you'll

19:25

notice is we have this little green

19:27

icon. So, each of our bots or agents we

19:30

have basically have their own icon. And

19:32

whenever it's moving like this, you

19:34

could see on the left-h hand side, this

19:35

basically just means that it's working

19:37

in real time. If it is stagnant, like

19:39

some of these ones right here, that

19:40

means that it's not performing any

19:42

tasks. I think this is really cool to

19:43

kind of just see which of the agents are

19:45

working at any given time. All right, so

19:47

just a few minutes later, it actually

19:48

went ahead generated this invoice in the

19:50

exact style. This is great. We can go

19:52

ahead and download this directly onto

19:53

our computer. Let me just go ahead

19:55

download that. And then now it's

19:56

basically saying that this is done. Say

19:58

the word when you want to generate the

20:00

next invoice or want me to pull what's

20:02

owed from ClickUp. So since this is

20:04

already connected to ClickUp, it's able

20:06

to go and pull the other invoices that

20:09

I've either sent or generated. So I

20:11

could see basically what needs to be

20:13

paid and what has been paid. Now, this

20:15

is really cool. We obviously have an

20:17

agent that has a specific style of

20:18

invoices that it's able to generate. But

20:20

where this gets really interesting is

20:22

when we come over here to create a

20:25

routine. This is where we could either

20:27

have this run on a schedule where every

20:29

single day this invoice agent will go

20:31

and scrape from my email inbox and then

20:33

identify what invoices we need to

20:35

generate, but also I could click on

20:37

create invoice and we can add a specific

20:40

trigger. So, what I mean by this is we

20:42

could basically connect this to a Slack

20:44

channel. So whenever there is a message

20:46

related to invoices, it will

20:47

automatically trigger this agent, which

20:50

this is the first time we've kind of

20:51

seen this from these AI agent platforms.

20:53

Before we could do scheduled tasks and

20:55

whatnot in Cloud Co-work, but we didn't

20:57

really have the option to allow

20:59

automations to kick off from a specific

21:01

trigger or event that's happening inside

21:03

of our apps. Now, I want to show you

21:04

something really cool. Now, if I come

21:06

over to my inbox triage agent, you can

21:09

see there is this conversation happening

21:11

between the inbox agent and the invoice

21:14

agent. And I didn't even have to go and

21:16

tell it to go and do this. It just

21:17

automatically knew intuitively that it

21:19

needed to communicate with other agent

21:21

in order to get this done. So, invoice

21:23

agent said, "Hey, Brock set me up as the

21:25

agent. Please hand me anything invoice

21:27

or payment related from our email

21:29

triage." So, it basically broke down

21:31

everything that it needs to flag for us.

21:33

And then the inbox triage agent went and

21:35

pulled through my email and found the

21:37

specific things that I need to, you

21:39

know, understand and be up to date with.

21:41

All right. So, what I'm going to do now

21:42

is create a group chat with the invoice

21:45

agent in my inbox agent. And what I'm

21:48

going to do is I'm just going to give it

21:49

a brain dump, but say, "Hey, I want you

21:50

to go and check my emails from the past

21:52

day. Find any email that is invoice

21:55

related. If there are specific invoices

21:57

that we need to create, use the invoice

21:59

agent in order to generate the invoices.

22:01

Then I want you to draft the reply with

22:04

the invoice in the email as well as I

22:07

want you to then add these to ClickUp in

22:09

order to track the process and the

22:12

status of all of these different

22:13

payments. Now, to show you real quickly

22:15

inside of my email inbox, I went ahead

22:16

and drafted three different emails as

22:19

demos because I don't want to show like

22:20

actual invoices that I need to create

22:22

just for confidential reasons. What it

22:24

did is it pulled from my email inbox and

22:26

now you could see basically all of the

22:28

invoices that we need to generate right

22:30

here. So, what it's going to do is now

22:31

generate the invoices, create the draft

22:34

reply for the email, so I could just go

22:35

and send it myself. And it's then going

22:37

to update this inside of ClickUp so I

22:38

could see the status of all these. All

22:40

right. And just like that, we now have

22:41

these emails drafted. So, for example,

22:43

if I click on this one, you can see that

22:45

it added this reply email as well as

22:47

even attached the specific invoice. And

22:49

I could go ahead and pull this up right

22:50

here in the exact style that we fed it.

22:52

as well as if I come into my invoice and

22:55

payments board inside of ClickUp, we can

22:57

now see that all of these invoices have

23:00

been submitted. As well as if I click on

23:02

this, it shows the client information

23:04

here, shows the email thread, and now me

23:06

and my team can see exactly where we're

23:08

at in the invoicing process. So guys,

23:10

there we have it. That is Grockbot, how

23:12

it works, a bunch of different use cases

23:13

that you can use, and it is now

23:15

available for free for you to try out or

23:17

you could have access to it on the $60

23:19

per month plan. Like I said, I've been

23:21

using this every single day since this

23:22

has released, and it's completely

23:24

changed how I've been using AI. I've

23:26

almost completely switched from using

23:27

Claude Co-work to now just texting

23:29

Grockbot on my phone all throughout the

23:31

day in order to get different tasks done

23:33

with my different agents I've

23:34

configured. And it's really powerful if

23:35

you connect it to the specific apps you

23:37

work across. So, if you guys got some

23:39

value from this video, leave a like,

23:40

subscribe to this channel, and let me

23:42

know what you think of Grockbot. And if

23:43

you want more videos like this, just

23:44

make sure to subscribe. And I have

23:45

plenty of content coming out soon.

23:47

Thanks for saying the end and I'll see

23:48

you in the next

Interactive Summary

This video explores the capabilities of Grockbot, an AI agent platform that has significantly improved the creator's productivity by replacing older tools like Claude Co-work. The presenter demonstrates how to use multiple agents, integrate various third-party applications via plugins and Zapier, and automate complex workflows such as managing emails, tracking meeting notes, and generating invoices.

Suggested questions

4 ready-made prompts