HomeVideos

I Built the Ultimate Team of AI Agents in 20 Minutes (Grok Bot)

Now Playing

I Built the Ultimate Team of AI Agents in 20 Minutes (Grok Bot)

Transcript

607 segments

0:00

In this video, I'm going to be showing

0:01

you how to build the ultimate army of AI

0:03

employees for your business using the

0:05

latest Grok bot. If you didn't know,

0:08

Grok bot is the latest AI product by

0:10

SpaceX or and xAI, and they are

0:13

basically marketed as AI teammates that

0:15

can do work and not just talk to it like

0:17

Claude or ChatGPT, but they can actually

0:19

sign into your tools like email,

0:21

calendar, CRMs, and use the tools just

0:24

like you do. And they can even learn how

0:26

you do tasks. For example, you can

0:27

screen share and teach Grok bot how to

0:31

do a certain task, and it will learn

0:33

from that screen sharing just like a

0:34

real human would and perform that task

0:37

for you. So, in this video, I'll be

0:39

building six different AI agents

0:41

together to form an AI army of

0:44

employees. And if you don't know who I

0:45

am, my name is Adam Chan. I've been

0:46

running my own AI agency Legacy AI over

0:48

the past 2 years. And if you're a

0:50

business owner you want to work with me

0:51

and my team, you can use the link in the

0:52

description to book a free discovery

0:54

call with us. First, let me first show

0:56

you what we are actually building today.

0:58

So, we are building an an AI team of

1:00

employees. First of all, we have our

1:02

first agent, which is the chief of staff

1:04

agent, and it connects to nothing. Its

1:06

whole purpose is just to route work and

1:08

never does the work. So, they would be

1:10

basically leading a team of different

1:11

bots and telling them, "Okay, for this

1:14

specific task, which bot would be most

1:16

suitable for that task?" And for the

1:18

actual employee, the AI employee, there

1:21

are five different bots. The first one

1:23

is an email or inbox bot. All it does is

1:26

basically connect to Gmail or Outlook

1:28

and check for the emails and reply if

1:30

necessary or ignore. The second bot is a

1:32

calendar bot, which is used for

1:33

appointment bookings. The third one is

1:35

prospect bot. This one actually does

1:37

research over prospect over the

1:39

internet. I'll be using screen share,

1:41

and that bot is actually connected to my

1:43

browser or computer use, so that it

1:45

would actually do the research on its

1:47

own. The fourth one is the content bot.

1:49

And the content bot is used to draft

1:50

well any content for me. And we have the

1:54

ops bot, which is actually connected to

1:56

my CRM, and then it can check what the

1:58

deal status is and what's missing, etc.

2:01

So, the first step that you want to do

2:02

is actually go to xai.bot. So, you can

2:04

see meet Grok Bot. So, what you want to

2:06

do is download for macOS. So, I'm just

2:07

going to click download. And after you

2:09

download, you should have an app that

2:11

look calls Grok Bot. Once you open it,

2:13

you'll be asked you to sign in to your

2:14

account, and you can either do the xAI

2:17

subscription or a Cursor account. If you

2:19

don't know what Cursor is, Cursor is

2:21

basically an IDE, so basically where we

2:24

can actually code just like Visual

2:26

Studio Code, right? It's a development

2:28

environment. And one of their plan is

2:31

that with Grok Bot, you will need a

2:32

Cursor account. And you can either get

2:34

the one-week trial, or you can get the

2:36

ultra which is $200 a month. So, for

2:38

now, we're just going to be using trial,

2:40

and I'll just click continue. And I'll

2:41

just say sign in to Grok Bot. Just click

2:43

sign in. Open Grok Bot. I'll just click

2:45

sign in. Sign in. Open Grok Bot. And

2:48

now, we have the Grok Bot set up

2:50

complete. I'll just click next. I'll

2:52

just click next. Next. What do you use

2:54

every day? So, I'm just going to click

2:55

on Google Workspace, Slack, Notion,

2:58

LinkedIn. Look, I use GitHub, not really

3:00

the rest. I use Calendly. I do use

3:03

Calendly, Loom. Yeah, that's it for now.

3:05

Again, feel free to adjust it as you

3:07

need it for your business. I click next.

3:09

And now, we have the entire interface

3:12

for Grok Bot. So, the first thing you

3:14

want to do is what we want to create a

3:16

bot. So, I'm going to be creating the

3:18

first bot which is kind of the chief of

3:19

staff, right? So, I'm just going to

3:21

click on chief of staff. And I'll click

3:23

start it. And then it says, "Hey, I'm

3:25

your chief of staff. Here to take the

3:26

operation load so you can stay on what

3:28

they actually need to." Right. So, now,

3:30

I'm going to give it a prompt so that it

3:32

will actually know what it's actually

3:34

doing. So, I would say, "Your days are

3:35

already running through Slack, Notion,

3:37

GitHub, and Google." I can plug into

3:38

this now so I'm not saying blind. I want

3:40

you to connect this now. I'd rather just

3:41

tell me the first job. So, I'll just say

3:44

here, I'll just kind of copy the prompt.

3:46

I'll just say, "You are my chief of

3:48

staff. You manage a team of five

3:49

specialist bots, and you're my single

3:51

point of contact. I talk to you not to

3:52

them. So, my team will be have the inbox

3:55

spot, which is the email, calendar bot

3:57

for scheduling, prospect bot for

3:58

research on companies and people, the

4:00

content bot for turning raw material

4:01

into posting drafts, and our sales the

4:03

sales spotlight. Um so, read what I

4:05

asked on this side which specialist are

4:06

needed, and then if you need some more

4:08

than one, work out the order, research

4:09

before writing, facts before I reach,

4:11

and never ask two bots for the same

4:13

thing.

4:14

And then third one is delegate with

4:15

specific self-contained instruction.

4:17

Give the specialist everything it needs.

4:19

It can only see our conversations. And

4:20

then wait for them. If one comes back

4:22

saying they're wrong, send it back with

4:23

what is missing. Do not accept a fake

4:25

answer and pass it to me.

4:27

And then the fourth one is come back to

4:28

me with one answer, not a transcript,

4:30

not six reports.

4:31

And then we'll just have a kind of work

4:32

session. So, read what I asked um and

4:35

then just try to do the research before

4:36

writing and in delegation. Basically, a

4:38

lot of what I repeated. I'll just say I

4:40

run um

4:41

for context, I'll just say I run Legacy

4:43

AI.

4:44

Just click send. And as you can see, it

4:46

automatically create the team of the

4:48

bots that I needed needed to create,

4:51

right? I run them and you get one

4:52

answer, right? I'm the single point of

4:54

contact building the five specialists

4:55

now. Teams up, inbox, calendar,

4:57

prospect, content, and ops. You talk to

4:58

me, I run them, you get one answer. Send

5:00

the first job, right? So, let me just

5:01

check

5:05

um let's see the inbox bot description

5:07

in here. So, you say, "You are inbox

5:09

bot, the email specialist on Legacy AI

5:11

chief of staff team. You handle the

5:13

email triage, drafting, and follow-ups.

5:14

You report only to chief of staff, never

5:16

to the user directly. When given a task,

5:18

do it fully and return one clear

5:19

result." What you found, what you

5:20

drafted, the sands, etc. So, by just

5:23

messaging the chief of staff itself

5:25

configured the um inbox bot, right? So,

5:29

you can either like By the way, each of

5:31

these bots is have its own virtual

5:33

computer. So, it's kind of working Think

5:36

of it like a human, right? So, if you

5:37

hired a administrative um instructor or

5:40

an admin, they would have their own

5:41

computer. Similarly, each of these bots

5:44

would have their own virtual computer

5:46

they can work on. So, if I open the

5:47

screen

5:49

every second now. So, and see that this

5:50

is the screen that they're using. So,

5:51

they have their own like Google Chrome

5:53

to to search for for example, I'll just

5:55

choose weather today.

5:58

And it will say like the weather today,

5:59

right? This fine. They have their own

6:00

terminal

6:02

and they have their own file manager to

6:03

be using, right?

6:05

So, I can just let me just close this.

6:07

Close this. So, for the inbox box, I

6:09

actually want to connect it to Gmail.

6:12

So, we're saying installing Gmail now.

6:14

I'm just going to authorize.

6:17

You can close this tab. All right. So,

6:18

now let me just go to Grok bots.

6:21

Start I'll just say it's started.

6:23

So, it will just say you can see like

6:24

Gmail is connected pulling the inbox now

6:26

so I can send Chief of Staff a first try

6:29

it, right? So, it's connected to the

6:31

user Gmail.

6:33

So, it said messaged Chief of Staff,

6:35

right? So, it messaged the Chief of

6:36

Staff. Gmail is live etc. You can see

6:38

like this is

6:39

some emails that we have. Fairly charged

6:41

here. This is I forgot to pay my quota

6:43

subscription. Sorry, guys. But, anyway,

6:45

you can see from the Chief of Staff bot,

6:48

it will say got it. I'm the single point

6:50

of contact during the first five such

6:52

that teams operate. And then you get the

6:53

message from inbox bot. So, now the

6:55

Chief of Staff got the message from the

6:57

inbox bot and is now telling me what to

6:59

do. So, this is the advantage of having

7:01

like the the the the first layer which

7:03

is kind of the Chief of Staff bot. The

7:06

reason why is because

7:07

I if I have like a hundred of these

7:09

bots, I only need to talk to one bot,

7:12

not a hundred of these bots, right?

7:14

So, it will say okay, this is the email

7:16

that I'm getting.

7:17

So, it will say want me to open an

7:18

Arabic building so you can update that

7:19

card. So, I'll just say not now because

7:21

I've already done it. But, you can just

7:22

like to it will automatically

7:24

turn on turn on tasks

7:26

exactly that you need to do and it will

7:28

inform you. So, now I can just ignore

7:30

it. And now let me configure it the rest

7:32

of the bots. So, let's turn on calendar

7:34

bots. So, it will say hey, I'll take

7:36

scheduling, availability, meeting prep

7:38

from Chief of Staff and gets them done.

7:40

I'm checking whether the is successfully

7:41

set up so I can start, right? So, I'll

7:43

just click on connect to Google

7:45

Calendar. So, I'll just click on

7:46

authorize.

7:47

I'll just select all and click continue.

7:50

I'm going to click close this app. So,

7:52

you get two messages with Chief of

7:53

Staff. So, you can see the Chief of

7:54

Staff is communicating with the calendar

7:57

bot, right? So, uh let me just see the

7:59

Chief of Staff.

8:00

Um

8:02

Calendar bot's connect us. You know,

8:03

they report to me

8:05

with the Google Calendar when there's an

8:06

actual scheduling job. Nice. Uh we're

8:08

waiting for scheduling work from the

8:10

Chief of Staff, right? Nice.

8:12

So, now we have the calendar bot

8:14

connected. Um let's do the Let's do the

8:16

prospect bot bot first, because the

8:19

prospect bot I have to actually teach

8:20

them how to research a lead, right? I'll

8:23

just say, "I'm going to be sharing my

8:24

screen to train you how to research a

8:27

prospect properly, okay?" I'll walk me

8:29

through it. I'll watch you how you

8:30

actually research a prospect and lock

8:31

that in, right? So, I'm going to be

8:34

clicking on the prospect bot screen. I'm

8:36

going to open it, and I'm going to click

8:38

on teach a task. And now, this whole

8:40

screen is actually recorded for the

8:42

purposes of training that prospect bot

8:46

to do exactly what I teach you to do.

8:48

So, first of all, I'll just click on

8:49

Google Chrome. Let's say I have to

8:51

research a prospect. Uh let's say I'm

8:53

researching plumbing company right to

8:55

work with. And I'll just say, I'll just

8:57

type in plumbing company. Let's say we

8:59

just want the phone numbers and we're

9:00

doing cold calling um for these

9:02

companies, right? So, then you would

9:04

just go to like Rain Plumbing Services.

9:07

You go there, and then you will try to

9:08

find their phone numbers. You've got the

9:10

phone number, and we're doing cold

9:12

calling for these clients, right? And

9:14

maybe I just need to know like the name

9:15

of the businesses. So, if I go to the

9:16

website, maybe I want to I want the the

9:19

agent to research the rainplumbing.com,

9:22

the website, the phone number. And here

9:23

you go, there's an email,

9:24

service@rainplumbing.com.

9:26

So, this is where I can outreach to

9:27

them, right? So, now I'm just teaching

9:30

the prospect bot how to look up the

9:33

information and save this information,

9:35

right? So, this is now we have the the

9:37

email, right? So, this is actually how

9:39

we train it. And now I just click finish

9:41

recording and there's just

9:43

like learn from demonstration. So, turn

9:46

on screen going to reusable skill. So,

9:47

this is actually managed by cursor. So,

9:49

this is actually a function. Think of it

9:51

like a template. So, learn from

9:52

demonstration its name the description

9:54

is is is basically a tool. Turn on the

9:56

screen recorded demonstration in

9:57

computer reusable skill. And then you

9:59

would basically learn from demonstration

10:01

you're giving a screen recording of user

10:02

demonstration and tasks on your computer

10:04

and then etc. So, this is basically

10:07

instructions by cursor that you don't

10:08

have to build yourself. And from this

10:11

screen recording the prospect will be

10:12

trying to learn what is it. So, he's

10:14

saying watch the recording now I'll put

10:16

the steps out and turn into skill.

10:17

You're on LinkedIn looking at a person

10:18

looks like Daniel at services firm and

10:20

the company page for people watch for

10:22

correction. You are on Crown Plumbing

10:24

sites. Then Google searching plumbing

10:26

company emails. Right, exactly what I'm

10:28

trying to do. Knockout Plumbing shows up

10:30

in the knowledge panel watching the walk

10:32

full walk through now, right? Still

10:33

watching the full clip. Apparently I'm

10:35

pulling the tabs and search here to see

10:36

the results from that session so the

10:38

URLs etc. By the way, while it's think

10:40

while it's talking you can also create a

10:41

routine task within GrokBot so that

10:44

these are recurring tasks this agent

10:45

runs on schedule. For example, I can

10:47

create a routine where we say research

10:49

prospect, right? So, I'll just say

10:51

research prospect find the emails and

10:53

phone numbers and the websites and the

10:55

company names of different plumbing

10:56

companies, 20 plumbing companies of 20

10:59

plumbing companies, right? And then when

11:01

to run so I can adjust on the schedule.

11:03

So, I'll just say every day at let's say

11:07

9:00 a.m. Right, let's say 9:00 a.m. So,

11:10

it will just say a every day at 9:00

11:12

a.m. it will try to find the emails and

11:14

phone numbers and websites and the

11:15

company name

11:16

um of 20 plumbing companies, right?

11:18

Exactly. And then I can just turn it

11:20

active and inactive, right? etc. So, now

11:22

it's seeing the pattern which is Google

11:24

a local trade open GPP websites. Uh hit

11:28

contact check maps the phone address and

11:30

booking link and then hunt emails for a

11:32

hunter rather than the bulk list. So, I

11:34

can just message it to correct it,

11:35

right? Based on his training. So,

11:37

actually what I want you to do as you

11:38

can see from the screen recording is go

11:40

to Chrome and and then look up uh some

11:44

plumbing companies and then from Google

11:46

Maps you can see its phone number and

11:47

the website name. And then from the

11:49

website name I want to go into there to

11:51

see if you can find an email for it.

11:53

Essentially, you are a uh lead getter

11:56

bot so that every morning you would run

11:58

the research prospect routine in order

12:00

to get me leads for plumbing companies

12:02

for my AI agencies skipping bulk

12:04

mailing. Now, this is pretty This is um

12:06

pretty good conclusion, guys. Yep, so

12:08

now proceed. This is maps for phone and

12:10

website then the site for email uh

12:12

plumbing unless you name in the city.

12:13

Grabbing a first batch now so you have

12:15

something tonight. Right, so it's

12:16

actually go ahead going ahead and

12:18

actually you can see what it's doing.

12:20

It's going to the plumbing companies for

12:22

San Jose, CA. Save the method as

12:25

research for first leads landing

12:26

shortly. Right, so now it's actually

12:29

scrolling through this list and it's

12:31

actually using its own virtual computer

12:34

and actually performing tasks that I

12:36

trained it to do, right? Pretty cool.

12:37

And now it actually went to the website

12:40

on its own, right? And doing this task.

12:42

So, now while it's doing this task,

12:44

while it's actually just like a human

12:45

employee, we'll look through the website

12:47

and try to get the phone number and the

12:49

leads information. And let me now go do

12:51

something else, content bot. So, yeah,

12:53

this is basically just drafting um

12:55

content, right? Um this is fine. And

12:58

then now I have the ops bot which is the

13:01

the operation bot. I'm on the pipeline

13:03

and deal hygiene from here checking what

13:05

real due dates I already have so I can

13:06

get a real status together.

13:09

So, I'll just say, "Can you connect to

13:10

Slack, please?" Because all of our

13:12

clients live in the Slack space. Slack

13:14

is the right place if that's where the

13:15

clients live. Checking the connector and

13:17

getting it connected, right? Now, I'll

13:19

just click on authorize.

13:20

Uh I'll just click on my workspace and

13:23

then just click allow and close this tab

13:26

and go back to your GrokBot. It's like

13:28

it's connected, right? And it will look

13:30

through the workspace next and start

13:31

mapping where the clients actually live,

13:33

right? So, now as you can see, the

13:35

prospect job finished researching

13:37

everything and it got eight San Jose

13:40

plumbers from the Google Maps and three

13:42

of them have public emails and they save

13:44

all of them in a nice CSV, right? So,

13:47

let me just go down here.

13:49

CSV, let me just save it and let me open

13:52

Google Sheets just to show you guys uh

13:55

um what it looks like. Well, I'll just

13:57

uh

13:57

open. So, you can see it got all the

13:59

name, it got all the phone number, it

14:01

got the website, the one that they

14:03

found, and the one that has emails, and

14:05

the one that has address, and the Google

14:06

Map links, and the notes. Like, this is

14:09

pretty good. Like, as a first run, this

14:10

is pretty cool. Like, it's very very

14:12

useful, right? As a lead getter,

14:14

essentially. So, you can just train it

14:16

one time and it can perform the task

14:18

that it needs to, right? And let's go

14:19

back to GrokBot. So, we say pipeline 10.

14:23

Um so, yep, booked today. I'm going to

14:25

be chased. So, these are basically all

14:27

the deal Well, all of my existing

14:29

client. And then, the operations bot,

14:31

yep, this is from the op So, this is

14:33

actually from the chief of staff bot.

14:34

So, the ops bot actually tell the chief

14:36

of staff bot like what's happening. So,

14:38

just say uh pipeline 10 etc. So, the ops

14:40

will actually actually summarize all the

14:42

channel status and actually put together

14:45

a markdown file of all the status of all

14:47

of our clients, right? So, now let me

14:49

just talk to chief of staff and ask it

14:52

to run through a diagnosis. I'll just

14:54

say uh "Hey, can you look through just

14:56

for today just to see what appointments

14:58

I have and just to see is there anything

15:01

urgent for my team to do right now? And

15:03

can you also ask like prospect bot to

15:06

also start gathering leads for me for,

15:09

let's say, a uh London-based uh

15:12

restaurant, right? For my AI agency

15:14

clients."

15:15

And please also help me draft together a

15:17

LinkedIn post to do today to do with AI

15:20

voice agent. And just check the emails

15:22

to make sure that I didn't miss anything

15:23

important today. This is a big task. So,

15:26

let's see if the chief of staff can

15:27

actually delegate a task correctly to

15:29

the appropriate sub agent to be doing

15:32

the things correctly. Today's calendar

15:34

that was urgent tonight today on the

15:35

email sweep, restaurant leads and the

15:37

voice agent LinkedIn draft, right? So,

15:39

he messages all these team members, the

15:42

the off spot and the calendar board and

15:43

the prospect bots are all kind of

15:45

getting their ping and they are then

15:47

they're already trying to figure it out

15:48

essentially. Okay, so it's messaged five

15:51

agents. So, all five agents has been

15:53

messaged essentially. Offset nothing's

15:55

on fire tonight. Still waiting on

15:56

today's calendar. So, offset you said

15:58

nothing. See, so you see the the off

15:59

spot didn't reply to me. The off spot

16:01

replied to the chief of staff, right?

16:03

Calendar is clear for today. Nothing's

16:05

booked tonight. That's correct. Inbox is

16:07

clean of important mail today. Cloud

16:08

building is the only one I've really

16:10

like paid for it. That's fine. LinkedIn

16:12

draft is in waiting on the restaurant

16:13

leads and you'll get the full pack,

16:15

right? So, the content bot actually

16:18

finishes drafting a LinkedIn post and

16:20

then go back to the chief of staff. By

16:21

the way, if you want to like make it

16:23

refine it further because as a content

16:26

bot itself, like the model is not very

16:28

good at like writing fully fully human

16:30

like post. So, what you can potentially

16:32

do is you can literally go on the screen

16:34

and then you can write a LinkedIn post

16:35

yourself or looked at viral LinkedIn

16:37

post. And from there the content bot can

16:40

then learn from what a good LinkedIn

16:42

post would look like and therefore write

16:43

better LinkedIn post, right? So, if you

16:45

want the bot to be performing at better

16:47

things, train it like how you train an

16:49

intern. Right? Just do it on inside the

16:51

computer screen. That's the beauty of

16:53

using Rock Bot. There's very little

16:55

prompting as you can see, but there's a

16:57

lot of there is some prompting, but

16:58

there's a lot of more just showing them

17:00

how to do it. And to be honest, the the

17:02

the core idea of to of the entire Rock

17:06

Bot, the beauty of it is just how you

17:07

organize your team, right? And the

17:10

prospect bot is working, right? It's

17:11

looking through the the internet, right?

17:14

Okay, so you got like the LinkedIn

17:16

draft. Okay, obviously this is not the

17:18

best post you've ever seen. The But to

17:21

be fair, it this is just from one

17:22

sentence, create a LinkedIn post. It

17:24

doesn't know what a good LinkedIn post

17:25

is. So, exactly, if you want to improve

17:26

upon it, do the method I just showed

17:28

you. Or better, if you literally just

17:30

upload like files of like past viral

17:33

LinkedIn posts, if you want to. Now,

17:34

this basically said 10 restaurant leads

17:36

to average strongest flat iron dish room

17:39

full list with resources on my computer

17:41

at prospects London restaurants or MD.

17:43

So, basically, it's saved a markdown

17:45

file of different prospects in the

17:47

prospect book, right? And then it got

17:49

more independent just landed on the same

17:50

list. Strongest free at still file.

17:52

Okay, cool. Okay, can you I'll just say,

17:55

"Okay, can you send me the full CSV

17:57

file, please?"

17:59

Because technically, you can go to the

18:01

computer and then you can go to like the

18:03

file manager and you can see the

18:06

prospects London restaurants markdown.

18:11

I'll just say I'll just say open. Okay,

18:13

so it

18:14

looked at the markdown file is say the

18:16

client,

18:18

the area, website, the sources, right?

18:21

All the different notes they research

18:23

over its virtual computer and then just

18:25

going to close this task. Close this.

18:28

Close this.

18:29

Going to go back.

18:31

And now go to the chief of staff making

18:33

the CSV. So, now you've created a CSV as

18:35

well that I can download. So, you can

18:36

see this is in an Excel format, the

18:38

name, restaurant name, area, website,

18:42

why public contact, sources, page fit,

18:45

etc. Really nice. Yeah, so this is

18:46

pretty good from Joseph Bonnici example.

18:48

So, this is actually how you would be

18:50

able to build

18:51

an AI employee for your business. And if

18:54

you enjoy this kind of content, remember

18:56

to subscribe. And based on my

18:58

background, this is something that I

18:59

like to do at the end of every video.

19:00

So, if you're new to this channel, you

19:02

might not notice, but at the end of the

19:03

video, I would just I would just love to

19:05

pray with

19:06

people who are watching this video. So,

19:07

if you're comfortable with that, stay to

19:08

the end of the video. If not, thank you

19:10

so much for watching and I hopefully

19:11

I'll catch you in the next one.

19:13

So, let us pray. In the name of the

19:14

Father, and of the Son, and of the Holy

19:15

Spirit. Amen.

19:17

Heavenly Father, thank you very much

19:18

again for giving me a platform to

19:20

educate the public and everyone else.

19:23

And I hope everyone who's watching the

19:24

video have a blessed day ahead of them.

19:26

Through Christ our Lord. Amen. In the

19:28

name of the Father, and of the Son, and

19:30

of the Holy Spirit. Amen.

19:33

Thank you and I'll see you in the next

19:34

one.

Interactive Summary

This video provides a comprehensive guide on building a team of AI employees using the xAI Grok Bot. The creator, Adam Chan, demonstrates how to establish a hierarchy consisting of a 'Chief of Staff' agent that delegates tasks to five specialized agents (Inbox, Calendar, Prospect, Content, and Ops). He explains how to set up each bot, connect them to relevant business tools like Gmail, Google Calendar, and Slack, and train them through screen sharing and demonstration. The video showcases how these agents interact to perform real-world tasks such as prospect research, email triage, and content generation, emphasizing the efficiency gained by having a centralized AI management system.

Suggested questions

3 ready-made prompts