This NEW AI Agent Creation Platform Will Blow Your Mind! BYE GPTS...

Liam Ottley
21 Mar 202419:30

Summary

TLDR视频介绍了一个名为Zapia Central的创新AI平台,它结合了GPT的对话能力和自动化工作流程,使用户无需编码即可创建强大的AI代理。通过实时数据连接、行为定制和与6000多个应用的集成,Zapia Central为企业提供了一个集中的AI工作空间,尽管平台在对话理解和速度方面仍有待提高,但它为AI自动化领域带来了新的可能性。

Takeaways

  • 🚀 介绍了一个名为Zapia Central的新兴AI平台,它可能是GPT生态系统的一个有力竞争者。
  • 🤖 Zapia Central结合了类似GPT的对话助手和自动化工作流,提供了一个集中化的功能。
  • 🔍 Zapia Central允许用户通过AI助手与数据源(如Google Sheets、Google Docs和Notion)进行交互。
  • 📊 该平台支持实时数据的动态更新,使得AI助手可以访问和使用最新的信息。
  • 📈 通过Zapia Central,用户可以无需编写代码即可创建自定义的AI行为和自动化任务。
  • 💡 Zapia Central提供了一个集成的环境,可以连接超过6000个不同的应用程序。
  • 🌟 该平台的愿景是成为自动化进化的下一步,使用户能够通过友好的AI助手自动化更多的工作。
  • 🛠️ 尽管Zapia Central在某些方面仍有待改进,如对话理解和响应速度,但它为AI自动化提供了新的用例。
  • 🔧 通过Zapia Central,用户可以创建自动化后台工作流程,例如通过Slack消息或Google表格行的添加来触发操作。
  • 🎯 Zapia Central可能是AI自动化领域的一个重要参与者,特别是对于初学者和希望简化AI集成过程的用户。
  • 📚 视频作者推荐观众尝试Zapia Central,并在自己的业务中探索其潜力,同时提供了加入其AI自动化社区的机会。

Q & A

  • Zapia Central是什么?

    -Zapia Central是一个新兴的AI平台,它结合了类似GPT的对话助手功能和自动化工作流,允许用户通过友好的AI助手与各种应用进行交互和自动化操作。

  • Zapia Central与GPT的主要区别是什么?

    -Zapia Central不仅具备GPT的对话功能,还整合了自动化流程和工作流自动化步骤,提供了一个集中化平台来管理所有的Zapier集成和Zaps。

  • Zapia Central如何帮助企业采用AI?

    -Zapia Central通过提供一个易于使用的对话层,使得企业能够利用成千上万的可能性来创建与超过6000个不同应用的集成,从而简化和增强企业的AI采用。

  • Zapia Central如何实现与实时数据的连接?

    -Zapia Central允许用户连接实时数据源,如Google Sheets、Google Docs和Notion,使得AI助手可以访问和使用这些更新的数据,为用户提供动态的、最新的信息。

  • 在Zapia Central中,如何创建自定义行为?

    -用户可以通过点击创建新行为,设置触发器,提供指令和配置行为来创建自定义行为,而无需编写任何代码。

  • Zapia Central在自动化方面有哪些优势?

    -Zapia Central的优势在于它将GPT风格的对话与后台自动化结合,允许用户通过AI助手来管理和触发各种自动化任务,而不需要进行复杂的编码工作。

  • Zapia Central目前存在哪些局限性?

    -Zapia Central在对话方面处理线程和回答问题的正确性还有待提高,速度方面也有改进空间,并且目前还无法像Chat GPT那样使用代码解释器来生成图表和可视化数据。

  • Zapia Central适合哪些人群使用?

    -Zapia Central适合那些希望在没有编码背景的情况下创建和使用AI代理的人,包括AI代理商、自由职业者以及想要利用AI改善业务的企业家。

  • Zapia Central如何促进AI代理的民主化?

    -Zapia Central通过提供一个无需编码、直接连接Zapier应用和动作的无缝集成平台,使得人们可以更容易地创建和使用AI代理,从而推动了AI技术的普及和民主化。

  • Zapia Central的未来发展有哪些可能的方向?

    -Zapia Central可能会继续扩展其生态系统,包括Zapier表格、Zapier接口和Zapier画布等,同时可能会增加数据可视化功能,以帮助用户更好地理解和分析数据。

  • 如何开始使用Zapia Central?

    -用户可以通过访问Zapier Central的网站,创建账户并开始尝试构建自己的AI代理。通过实际动手尝试和解决自己生活中的问题,用户可以更好地学习和掌握这个平台。

Outlines

00:00

🤖 探索Zapia Central:AI自动化的新篇章

本段介绍了Zapia Central,一个新兴的AI平台,它是对Open AI的GPTs的有力竞争者。Zapia Central结合了GPT的对话式交互和自动化工作流,提供了一个无需编码的AI自动化解决方案。视频制作者Liam通过亲身体验,分享了Zapia Central的独特之处,包括与AI机器人的协作、使用实时数据源、以及创建自定义行为和触发器等。此外,他还强调了Zapia Central如何帮助用户提升业务,并且不需要任何编程技能。

05:01

📊 利用Zapia Central与电子表格和CSV数据进行交云

这一部分展示了Zapia Central如何与Google Sheets等电子表格工具集成,实现动态数据查询和分析。Liam通过一个示例,展示了如何从大量数据集中提取和比较2003年和2004年的销售数据。他还解释了Zapia Central如何将电子表格转换为数据库,并执行SQL查询,从而提供实时更新的数据和深入的业务洞察。此外,他还探讨了Zapia Central与Google Docs和Notion等文档工具的集成,以及如何通过对话式界面动态更新和查询文档内容。

10:02

🔧 Zapia Central的自动化和后台集成

在这一段中,Liam深入探讨了Zapia Central的自动化功能,包括如何设置基于特定短语触发的自动化行为,例如创建新的Notion页面。他通过实际演示,展示了如何配置和测试这些行为,以及如何将电子表格数据自动填充到Notion页面中。此外,他还讨论了Zapia Central如何与传统的Zapier自动化集成,例如通过监控电子表格的更新来触发Slack消息。Liam强调了Zapia Central作为一个集中管理平台的潜力,以及它如何帮助用户更有效地管理和自动化业务流程。

15:04

🌐 Zapia Central在AI自动化领域的地位和应用

最后一段讨论了Zapia Central在AI自动化市场中的潜在作用和应用。Liam指出,Zapia Central通过消除编码需求,使得创建AI代理变得更加容易,这可能会推动AI代理的民主化。他还提到了Zapia生态系统的快速发展,包括Zapia Tables和Zapier Interfaces等工具,以及它们如何为用户提供一个全面的AI解决方案。Liam鼓励观众尝试Zapia Central,并在自己的业务中寻找应用场景。他还提到了自己的免费学校社区,那里将提供更多关于Zapia Central的深入讨论和资源。

Mindmap

Keywords

💡zapia Central

zapia Central是一种新兴的AI平台,它结合了GPT风格的对话助手和自动化工作流程的功能。在视频中,它被描述为可能改变企业采用AI方式的下一代技术。它允许用户通过友好的AI助手与各种应用程序进行交互,实现自动化任务。

💡AI自动化

AI自动化是指使用人工智能技术来自动执行任务或流程,减少人工干预,提高效率和准确性。在视频中,AI自动化是通过zapia Central平台实现的,它可以自动化工作流程,如数据分析、文档管理等。

💡对话助手

对话助手是一种AI技术,能够与人类进行自然语言对话,提供信息、解答问题或执行命令。在视频中,zapia Central的对话助手可以与用户互动,根据用户的指令执行特定的自动化任务。

💡无代码

无代码指的是用户在创建应用程序或自动化流程时不需要编写任何代码,通过图形界面和预设功能即可完成。视频中强调zapia Central平台的无代码特性,使得即使没有编程技能的用户也能轻松创建和部署AI代理。

💡实时数据

实时数据是指即时更新的数据,能够反映最新的情况或状态。在视频中,zapia Central可以连接到实时更新的数据源,如Google Sheets,确保AI助手使用的数据是最新的。

💡集成

集成是指将多个系统、应用程序或服务连接在一起,以便它们可以协同工作。在视频中,zapia Central通过集成各种应用程序和服务,如Google Docs、Notion等,提供了一个统一的AI工作平台。

💡工作流程自动化

工作流程自动化是指使用技术自动执行一系列预定的任务或流程,以提高效率和减少错误。在视频中,通过zapia Central,用户可以设置自动化的工作流程,如自动发送消息、创建文档等。

💡知识库

知识库是一个信息存储系统,用于保存和管理组织的知识、文档和数据。在视频中,通过zapia Central,可以将实时更新的数据源作为知识库,供AI助手查询和分析。

💡触发器

触发器是一种用于启动某个动作或流程的条件或事件。在视频中,zapia Central允许用户设置触发器,当特定条件满足时,自动执行预定义的任务。

💡行为(Behavior)

行为在zapia Central中指的是用户为AI助手定义的一系列指令和动作,用于指导AI如何响应特定的触发器或查询。

Highlights

介绍了一个名为Zapia Central的突破性新AI平台,它是AI自动化的下一个大事件。

Zapia Central结合了GPT的对话能力和自定义动作的自动化功能,提供了一种独特的AI使用方式。

Zapia Central允许用户通过与AI机器人对话来自动化工作,无需编写代码。

Zapia Central支持实时数据连接,如Google Sheets和Google Docs,使数据动态更新。

Zapia Central可以查询大型数据集并提供有价值的业务洞察。

Zapia Central能够将电子表格转换为数据库,并执行SQL查询。

Zapia Central提供了一个集中化的AI工作空间,可以将所有自动化和AI功能集成到一个中心。

Zapia Central允许用户创建行为,这些行为可以响应特定的触发器并执行自定义动作。

Zapia Central支持与Notion等应用的动态数据连接,但可能存在一些需要改进的bug。

Zapia Central可以自动化创建Notion页面,并根据用户的输入生成内容。

Zapia Central可以作为学校社区管理的中心枢纽,自动发送Slack消息和执行其他自动化任务。

Zapia Central的AI层可以连接所有这些自动化任务,无需直接的输入匹配。

Zapia Central的不足之处在于对话方面的处理速度和上下文理解,还有待提高。

Zapia Central缺乏数据可视化功能,这是未来可能需要添加的功能。

Zapia Central可能会在AI自动化领域发挥关键作用,特别是在无需编码的情况下创建AI代理。

Zapia的生态系统正在快速发展,为想要进入AI领域的人提供了一个有吸引力的机会。

建议观众尝试使用Zapia Central,并在自己的业务中寻找使用案例。

Transcripts

play00:00

in this video we're exploring a

play00:01

groundbreaking new AI platform that no

play00:04

one is talking about meet zapia Central

play00:06

the next big thing in AI automation

play00:08

while recently all of the buzz has been

play00:10

about custom gpts zapia has been

play00:12

silently working away on their answer to

play00:15

open ai's agents and an answer that

play00:17

could mean trouble for the entire gpt's

play00:19

ecosystem but could zapia Central really

play00:21

be the future of how businesses adopt AI

play00:24

or is it just another stale automation

play00:26

platform that struggles to create real

play00:27

value for its users and more importantly

play00:29

if it's is any good how can you use it

play00:31

to boost your business or the businesses

play00:33

of your clients with AI well after

play00:35

diving into zapia Central over the past

play00:37

few days I have discovered that the

play00:38

zapia team is definitely onto something

play00:41

here so in this video I'll be breaking

play00:42

down what zapia central is and what

play00:44

makes it a potentially better

play00:45

alternative to open AIG gbts in many use

play00:48

cases then I'll be whooping up some live

play00:50

builds for you to demo some of these

play00:52

super simple yet stupidly powerful use

play00:54

cases that anyone can create on zapia

play00:56

Central and finally I'll be sharing with

play00:58

you how you can become one of the first

play00:59

experts on this new platform to boost

play01:01

both your business your client's

play01:03

businesses all regardless of your skill

play01:05

level if you're new to the channel and

play01:06

don't know who I am my name is Liam

play01:08

otley I run my own AI development

play01:10

company called morning site Ai and I

play01:11

also run the largest community of AI

play01:13

agency owners in the world but I teach

play01:15

people how to make money by selling AI

play01:17

solutions to businesses just like the

play01:19

agents and automations you can create on

play01:21

zapia Central let's get into it first

play01:22

things first what is zapia Central well

play01:24

here on this site for Central they say

play01:26

that you're able to work hand inand with

play01:28

AI Bots the easiest way way I found to

play01:30

explain it is a mixture between gpts

play01:32

where you're able to interact with a

play01:34

conversational assistant or an agent and

play01:36

tell it what to do and then it can call

play01:38

Custom actions and and use knowledge Etc

play01:40

so you've got the knowledge you've got

play01:41

the actions you've got the prompting

play01:42

usual GPT stuff that we've talked about

play01:44

on this channel quite a lot but what

play01:45

zappy has done is a very interesting

play01:47

aspect of also including some of the

play01:49

sort of autonomous Automation and

play01:51

workflow automation steps that they've

play01:53

done on the platform prior so it's a

play01:55

unique combination of gpts mixed with a

play01:57

centralization of all of your zaps for a

play02:00

particular function if that doesn't make

play02:01

sense don't worry it was difficult for

play02:03

me to WRA my head around this initially

play02:04

but give me a few minutes and once we

play02:05

get into the demos it's going to make a

play02:06

lot more sense scrolling down here a bit

play02:08

the vision that they have for Central is

play02:10

the next step towards evolving

play02:11

automation for the AI AG Central puts

play02:14

you in the driver's seat to automate

play02:15

even more work with the help of our

play02:17

friendly AI Bots what central really is

play02:19

is a conversational layer that allows

play02:21

you to utilize all of the thousands and

play02:23

millions of possibilities that you can

play02:24

create with zap GE zaps and the

play02:26

Integrations they have with over 6,000

play02:28

different apps so extremely exciting

play02:30

stuff here's a little bit of a demo you

play02:32

can connect live data so it's similar to

play02:34

knowledge for your gpts as you can see

play02:36

here add a data source there's Google

play02:37

Sheets as an option there's Google Docs

play02:39

and there's also notion the interesting

play02:41

thing about this that you're going to

play02:42

see in a second is that it is actually

play02:44

Dynamic so whatever the Google sheet is

play02:46

updated you're then going to have access

play02:47

to that updated data which is something

play02:49

we haven't really seen which is

play02:50

knowledge bases thata Dynamic and

play02:52

regularly updated to be update with the

play02:54

latest information live data is one part

play02:56

then we have the instructions or the

play02:58

prompting as we see here when they open

play03:00

this up there's an instructions tab

play03:01

which instructs your Bot on how to use

play03:03

the knowledge and also how to use the

play03:05

actions and finally they have what's

play03:06

called behaviors which are best

play03:08

understood as actions for custom gpts or

play03:11

tooling Etc this of course is directly

play03:13

integrated with all of zapia thousands

play03:15

of Integrations as you may have noticed

play03:17

these are the same three ingredients

play03:19

that I've been banging on about for the

play03:20

past couple of months and that if you

play03:22

can figure out this AI agent skill set

play03:25

which right now is knowing how to prompt

play03:26

knowing how to add data and knowing how

play03:28

to connect these tools the you're able

play03:30

to build AI agents on basically any of

play03:32

these platforms because it's the same

play03:33

skill set so I hope you've been paying

play03:34

attention and I hope you built up these

play03:36

skills of knowing how to prompt knowing

play03:37

how to connect different actions because

play03:39

it's the same process here tying all of

play03:41

these three aspects together into these

play03:43

AI agents that you can use internally in

play03:44

a business the biggest Advantage for

play03:46

zapia Central over gpts in my opinion

play03:49

especially for beginners is that there

play03:50

is no code included at all you don't

play03:53

have to add in these tools separately

play03:55

don't have to use a schema I'm sure if

play03:56

you've tried to get zap your actions to

play03:58

be put as custom actions for your gpts

play04:00

you'll know just how difficult that

play04:02

process can be so for this there is no

play04:04

code involved at all all you need to do

play04:05

is go on as it shows here click on

play04:07

create new Behavior give it some

play04:09

instructions set up a trigger and then

play04:11

set up what it's supposed to do and

play04:12

you're able to create these awesome

play04:14

powerful custom actions without ever

play04:16

having to write any code at all so down

play04:18

here they've got a little bit more stuff

play04:19

I think this allinone AI workspace is

play04:21

really the the key Point here is that it

play04:23

Con centralize all of your automations

play04:24

and AI features that you use throughout

play04:26

the business into one Hub which I guess

play04:27

is why they called it zapia central it's

play04:29

pretty pretty fitting title query data

play04:31

we been over that follow up these are

play04:32

the kind of backend automations you can

play04:33

set up that you'll see in a second Etc

play04:35

that hasn't got you excited about

play04:36

building a zapia central then the next

play04:38

section where we're going to jump into

play04:39

some live demos of building OD zappi

play04:41

Central will do the job so I can click

play04:44

on try Central today of course you need

play04:46

a zap your account for this once we're

play04:48

ENT Central we have this familiar kind

play04:50

of chat gbts interface and it

play04:53

immediately gives us a new Untitled bot

play04:55

and it tells us okay we can add

play04:57

behaviors we can add data sources a good

play04:59

starting point for us I think would be

play05:00

to show some of the live data and being

play05:02

able to chat to uh spreadsheets and and

play05:04

CSV data so if I add a data source go to

play05:06

a Google sheet it's connected by my

play05:08

email to all of my Google Sheets so

play05:10

super easy to do that now I've got this

play05:12

Sample Sales data that I've got off

play05:14

kaggle just to show you how even large

play05:16

data sets can be used and we can query

play05:17

our data sets to get some interesting

play05:19

insights to give you a quick look at the

play05:20

data uh there's a ton of stuff here

play05:22

there's the sales there's the quantity

play05:24

of orders there's the date most

play05:25

interestingly which I want to do a

play05:27

little bit of analysis over so see

play05:29

there's a lot of 2003 2004 dates let's

play05:32

give it a bit of a tricky question to

play05:33

analyze that information so I go uh can

play05:36

you get the sales by month for 2003

play05:38

include a number of units sold and the

play05:39

total value month by month and what it's

play05:41

doing here similar to chat GPT is that

play05:43

it has its own code interpreter so as

play05:45

you can see it's actually converted the

play05:46

spreadsheet into a database and it's

play05:48

writing a database query on this uh this

play05:51

is some really impressive stuff and it's

play05:53

being kind of a fantasy for a long time

play05:55

of being able to convert a spreadsheet

play05:57

into a database and now we're writing

play05:59

SQL queries going from text to SQL I'm

play06:02

sorry I'm getting a little bit nerdy

play06:03

here but it is very impressive what

play06:05

they've been able to do here with

play06:06

converting it to a database and now we

play06:07

can query the database and get some

play06:08

really powerful insights so here you can

play06:10

see the output of the data retrieved and

play06:12

then it's using that data to write the

play06:13

answer so here we can see uh the data

play06:16

for 2003 month by month unit sold and

play06:19

total sales value so very very cool

play06:21

stuff already now can say can you

play06:23

compare that month by month 2 2004 EEG

play06:25

Jan 2003 this jan4 now this capability

play06:28

alone to set for a business being able

play06:30

to chat to their spreadsheets in a way

play06:32

that's very Dynamic and can pull in

play06:34

literally Dynamic data as it gets added

play06:36

to the spreadsheet is an awesome new use

play06:38

case and I really recommend you guys

play06:39

jump in and have a play around with this

play06:40

because this is something that

play06:41

businesses have been wanting to do for a

play06:43

long time and having it put on a PL like

play06:45

this for us is some awesome news for

play06:46

beginners and Pros alike and here we can

play06:48

see again it's written a query for the

play06:49

database it's created for our

play06:50

spreadsheet it's done it perfectly it's

play06:52

comparing the 2003 values month by month

play06:55

with the 2004 value so Dynamic spree

play06:57

cheet datas added as a knowledge base is

play06:59

so much better than the knowledge bases

play07:01

we have for gpts I'm sure you have tried

play07:03

to put a CSV data or tried to put some

play07:05

spreadsheet data as a knowledge document

play07:08

for your gpts and it's just absolutely

play07:10

terrible at at using that data because

play07:12

it doesn't have a system setup like this

play07:13

to be able to create a database out of

play07:15

it and then query the database and use

play07:17

that intelligently in the answer that's

play07:18

how we can chat to spreadsheets with

play07:19

Dynamic data which is cool enough and

play07:22

then we have the typical chat to

play07:23

documents in the same way that you'd

play07:24

uploaded a document to a knowledge base

play07:26

but again it's going to be dynamic so if

play07:28

I go to Google Docs and I select a

play07:31

Google doc here if I go maybe the

play07:33

support documentation for ailm would be

play07:35

a pretty good one for this because if I

play07:36

update the support documentation I don't

play07:38

have to go in and redownload it and

play07:40

re-upload it into my agent it's just

play07:42

going to automatically do that so I can

play07:43

get my team to go in there modify the

play07:45

support documentation it's necessary add

play07:47

new question and answers and FAQs and

play07:49

it's going to automatically be pulled

play07:50

through to zapia central here

play07:51

interestingly this brings us to some of

play07:53

the quirks of this platform and its

play07:54

current form as you can see here I've

play07:56

uploaded the information I've said how

play07:58

do I change my ail and password which is

play07:59

included in the support documentation

play08:01

just in order to change your password

play08:04

however it's not developed by zapia and

play08:05

it just starts making it up based off

play08:07

its existing training date this standard

play08:09

chat interface is not very useful in

play08:10

many ways and in order to get it to do

play08:12

the things you want to do in many cases

play08:13

you'll have to go create behavior and

play08:15

say answer questions from the K Liam for

play08:18

document provided then if I go to

play08:20

trigger and go when I message the bot

play08:23

add the

play08:24

trigger turn that

play08:27

on and now if I go

play08:30

I for to my AI password finally after a

play08:32

pretty long wait I can click onto this

play08:34

thread and see that it has gone through

play08:36

and Ed the code block to select the

play08:38

document first of all and then to answer

play08:40

the question from the information in the

play08:41

document it doesn't quite work how you'd

play08:43

expect it in these simple use cases but

play08:44

I believe from my plan around this is

play08:47

set up this way so that if when you have

play08:49

tons and tons of documents like 20 or 30

play08:51

of them included this system of using a

play08:54

a look up and and looking up in the

play08:55

database to find the specific

play08:57

documentation file that you're looking

play08:58

for or the the the knowledge based

play09:00

document You' provided it's going to

play09:02

actually look that up first and then put

play09:03

it into this thread so when the actions

play09:05

are performed this is the interesting

play09:07

part where you combined a regular GPT

play09:08

Style with I think what Zapp is trying

play09:10

to do which is this threading style and

play09:12

so now I can conversate back and forth

play09:14

within this thread and ask questions

play09:15

about my document and I'm going to get

play09:17

the answers in there for dynamic

play09:18

knowledge bases you can also connect

play09:19

notion here I have had some issues with

play09:21

getting it to answer correctly so I

play09:23

think there might be a few bugs um but

play09:24

I'm sure they'll buff those out in no

play09:26

time here I've made a bot called my

play09:27

notion resistant and what I want to do

play09:29

is set up some behavior in this create a

play09:31

behavior instructions are a good place

play09:33

to start so I could say create a new

play09:35

notion page when I say create a new

play09:38

notion page ask me for the content for

play09:40

the page but come up with a page title

play09:42

yourself that represents the content

play09:44

appropriately and so now usually you'll

play09:46

be getting suggestions here I'm not sure

play09:48

why it's not suggesting me this can be a

play09:50

little bit confusing but just stick with

play09:51

me here so when I message the bot is the

play09:53

trigger then I want to use my phrase

play09:55

here oh which was create a new notion

play09:58

page so I can provide a specific keyword

play10:00

or phrase that I want to trigger this

play10:01

action create a new notion page I want

play10:04

to add that as a trigger and then I want

play10:06

to create a new page so here's my

play10:08

suggestions here that's nice the parent

play10:10

page have the AI select this so this is

play10:12

really the magic of of zapy central

play10:15

which you'll see a bit more on a bit but

play10:16

we have the title I want the AI to

play10:17

generate the field for the value and I

play10:18

also want it to cover with the content

play10:20

based off my input so that's all set up

play10:22

I can test the Behavior now it's going

play10:24

to open up a thread here to proceed with

play10:25

creating a notion page I need to know

play10:27

what content you'd like to include in

play10:28

the page here you can see it's triggered

play10:30

this is the test trigger that it's just

play10:31

inserted the phrase that I've included

play10:33

in the setup let's jump back to my

play10:35

spreadsheet chat one and I want to put

play10:37

this data into a new notion page see if

play10:40

it copies across okay so I to go back

play10:41

into the thread

play10:44

here oh let's see how it goes with that

play10:47

we got a little bit of a bug here where

play10:48

it didn't actually send me a reply

play10:49

normally it will send back the link to

play10:51

the notion page itself uh but I followed

play10:53

up and it seems like it's already

play10:54

created it so sure enough over on my

play10:56

notion it has created this it's given

play10:59

the title itself that it's generated

play11:00

based off the content and I've got an

play11:02

ocean page so maybe it was because it

play11:04

was in the testing environment if I

play11:05

tried again create a new noce page try

play11:07

it again without the testing environment

play11:09

now it's going to pick it up create

play11:10

another thread and see it started

play11:12

talking about some some random stuff

play11:14

here that's because I didn't come here

play11:15

and switch the action on that's my bad

play11:17

apologies and if I want to call this

play11:20

again and try to get it to properly

play11:22

create the notion page create a new

play11:23

thread for me hello seems to be bugging

play11:27

out here let me refresh the page Maybe

play11:29

little modification here I've just

play11:30

changed it to say trigger when I say

play11:32

create a new notion page so if I go

play11:34

create a new notion page there we go so

play11:37

make sure in this you use trigger it

play11:40

seems to be the keyword that gets it to

play11:41

work well so trigger when I say create a

play11:43

new notion page now it's going to be

play11:45

asking me in the thread for the content

play11:47

I want to include let's grab the sales

play11:49

data again head back to notion assistant

play11:51

go on to our thread and paste it in it

play11:53

should give me the confirmation message

play11:55

that it's created the page successfully

play11:57

propose page title now it's going to

play11:59

create the notion page please hold on

play12:01

for a moment and it's performing the

play12:02

action as expected there we go if we

play12:04

drop this down it has included all the

play12:06

information that it used to create it

play12:08

and it provides us a link to the report

play12:10

here so if I click on this can open up

play12:11

my ocean and there we have it strange

play12:13

thing is that it's kind of just putting

play12:15

it anywhere I think if I go in here it's

play12:18

down here and this one is inside the the

play12:20

other one I just did so back here if we

play12:22

look at the notion create page I can't

play12:24

even modify what the parent page is here

play12:26

so it's not perfect by any means but it

play12:28

does show you that this is a GPT style

play12:30

experience with these thread mechanics

play12:32

that allow you to set up certain

play12:34

behaviors to happen Upon A Certain

play12:35

phrase being sent this particular human

play12:37

in the loop style interaction with the

play12:39

assistants is not all zap Central does

play12:41

it's intended to combine both the GPT

play12:43

style where you are interacting with

play12:45

your agent with your typical background

play12:47

workflow automations that you set up on

play12:49

zapia which might be triggered by some

play12:50

action like a slack message or a Google

play12:52

sheet row added in so here's an example

play12:54

I've been working on just having a play

play12:56

around scratching my own itch with this

play12:57

stuff is a school control Center I have

play12:59

my free school community and I thought

play13:01

why to try to create some kind of

play13:02

central hub for managing it and put zapu

play13:04

Central to the test so I have a uh a

play13:07

link to a spreadsheet that is synced 8

play13:09

minutes ago so it's actively resyncing

play13:11

the data so I'm always able to ask

play13:12

questions about my school Community but

play13:14

more interestingly is I've been able to

play13:16

set up some Behavior where when someone

play13:18

tries to join my school Community a new

play13:20

robot added to the spreadsheet that it's

play13:22

watching and that is the trigger no

play13:24

longer am I having to ask a message to

play13:26

my assistant to trigger it it's able to

play13:28

use a traditional zapia trigger and then

play13:30

I'm getting it to send an action after

play13:31

that which is to send a direct message

play13:33

and the magic of zapia central is that

play13:35

the AI itself as I said is a layer on

play13:37

top that can connect all of these things

play13:38

so the instructions here will tell the

play13:40

AI how to autofill the rows so as you

play13:43

can see here for message text I need to

play13:45

generate some kind of message that I

play13:46

want to send to myself on slack but what

play13:48

I'm telling it bya instructions is when

play13:50

a new row is created send a slack

play13:52

message to me that is a one- sentence

play13:54

summary of the new members details if I

play13:56

go over to my slack here you'll see that

play13:57

zapia has been sending me messages and

play13:58

these one sentence summary obviously I'm

play14:00

going to have to blur these out but I

play14:02

promise you that there are one sentence

play14:03

summaries as I told it to over on zapia

play14:06

centroids triggering the automation

play14:07

based on a row being added and then it's

play14:09

using the AI in the instruction to

play14:11

determine how it fills out the rest of

play14:13

it so it's like this magical glue that

play14:15

can stick your automations without

play14:17

having to get this direct lining up and

play14:19

matching up of the inputs from one to

play14:21

another and as I said this combines the

play14:23

conversational aspect that we're used to

play14:24

with gbts with these kind of background

play14:26

and autonomous automations so if I go to

play14:29

while I was away we had 17 runs now I

play14:30

get to see everything that's been

play14:32

happening in the background of my school

play14:33

Community while I've been asleep while

play14:35

I've been away I can keep tabs on all of

play14:37

the automations that are firing I can

play14:39

set up new ones another cool thing you

play14:40

can do is go to create Behavior I can

play14:43

set a trigger to schedule so if I want

play14:46

to schedule something that I want done

play14:47

every week which might be to look at all

play14:49

of the spreadsheet RADS that have come

play14:50

in and all the new members that have

play14:52

tried to join and do some analysis and

play14:54

then send the analysis via email to

play14:56

whoever or to send it in here there's so

play14:58

much cool stuff that can be done now

play15:00

they we'll be able to combine both the

play15:01

human and the loop aspect of AI agents

play15:03

and talking to them and telling them

play15:04

what to do with these backend and and

play15:06

background automations that's really the

play15:08

magic of what zapia has done here with

play15:09

Central and I'm a big

play15:11

fan there are a few limitations that the

play15:14

platform suffering from I think the

play15:15

conversational aspect of pulling in

play15:17

context from the threads and answering

play15:18

questions correctly I think there's

play15:20

still a lot of work to be done there it

play15:22

is quite slow I've had to obviously cut

play15:24

it out for this video but some days I've

play15:25

been sitting here for 30 seconds or more

play15:27

waiting for it to to trigger in action

play15:29

or something like that there's

play15:30

definitely room for improvement on the

play15:31

speed side but this is quite clearly

play15:33

intended to be an internal use kind of

play15:35

platform and tool for managing your

play15:37

systems managing your operations

play15:38

managing the different departments and

play15:40

centralizing like the school example so

play15:42

centralizing everything to do with my

play15:43

free community all the automations all

play15:45

the reporting all the questions I want

play15:47

to have over the data I have really

play15:49

really cool way of centralizing things

play15:51

it's not so much as we've been familiar

play15:53

with with other agents of maybe creating

play15:55

a GPT and deploying it somewhere putting

play15:57

it out on an Instagram Channel or

play15:59

putting it on WhatsApp and things like

play16:00

this because it doesn't have to be

play16:02

interacting with customers I think the

play16:03

speed can kind of be forgiven if you're

play16:05

using it internally for now so not

play16:06

tearable but could definitely be better

play16:08

and one thing that would make this a lot

play16:09

more powerful is being able to visualize

play16:11

the data it's currently not able to like

play16:13

chat gbt use the code interpreter to

play16:15

generate a graph and so when I was

play16:17

looking at the data I would have lik to

play16:18

see the 2003 line against the 2004 line

play16:21

it's a lot more easy to understand and

play16:23

do data analysis when you can actually

play16:24

visualize it as well so I'm sure they've

play16:26

got that on the road map but it doesn't

play16:27

have that right now now that we know

play16:28

there zapia Central has got some pretty

play16:30

cool stuff going on the question is

play16:32

where does it fit into the the AI

play16:33

automation landscape how can we adopt it

play16:36

how can we include it in our own

play16:37

businesses how can we use it and sell it

play16:39

to our clients to me I think zapia

play16:41

central is going to play a key role in

play16:42

the space in a number of ways firstly it

play16:44

is completely remove the need to code in

play16:46

order to create these AI agents we used

play16:48

to have that integration step up okay

play16:50

have a a custom action or have an API

play16:52

that I want to put onto my GPT as a

play16:54

custom tool or a custom action and there

play16:56

was this ugly open API scheme thing that

play16:58

beginners had to do that wasn't very

play16:59

bitor friendly in many ways what zapia

play17:01

centralist enabled us to do is to

play17:03

directly hook into all of zapia apps and

play17:06

and the actions and triggers that you

play17:07

can use seamlessly no code not a single

play17:10

line of code has been written this could

play17:11

really be the democratization of AI

play17:13

agents in many ways that now people can

play17:15

come in without any coding background

play17:17

and they can start to build these things

play17:18

either for their own businesses or sell

play17:20

building them as a service to other

play17:21

people as well and when you factor in

play17:22

that zapia is making major moves on

play17:24

things like zapia tables for storing

play17:26

data they have zapier interfaces they

play17:28

have zap canvas as well their ecosystem

play17:30

is growing at a rapid rate and this

play17:32

means if you are someone who's looking

play17:33

to get into AI or building these kind of

play17:36

automation systems and beginning to make

play17:37

money by by making businesses better

play17:39

with them zapia may be a interesting

play17:42

opportunity for you to focus on just one

play17:44

ecosystem there is so much available in

play17:46

this AI space right now if you're an AI

play17:48

agency or you're a freelancer just

play17:49

picking one platform like zapia and

play17:51

saying look they've got so much stuff

play17:52

I'm just going to specialize in that

play17:54

that could be a great place to start

play17:56

especially if you're a beginner and you

play17:57

want to utilize no code tools as as much

play17:59

as possible so if what you've seen today

play18:00

seems interesting at all I highly

play18:02

recommend you go on to zap PA Central

play18:03

and check it out try to build just one

play18:05

thing and the best way to learn this at

play18:07

least in my experience is to scratch

play18:08

your own itch try to build something

play18:10

that's handy for you in your own life or

play18:11

for a friend or for your parents or

play18:13

something just try to find a a use case

play18:16

that you can apply this tool to help

play18:18

make your life better and that aha

play18:19

moment might just be the thread that you

play18:21

end up pulling to a particular career in

play18:23

this AI autovation space I'm super

play18:25

excited to see what kind of stuff you

play18:26

guys are building on zapia Central and

play18:28

being out of find better use case than I

play18:30

have so far so what I'm going to do is

play18:31

on my free school Community if you're

play18:32

not already a member there will be a

play18:33

link to join in the description I'm

play18:35

going to make a post related to this

play18:36

video that is on the topic of zapia

play18:38

Central and if you want to head over

play18:40

there we're going to be talking and

play18:41

going deep on all the cool features and

play18:42

use cases that we're able to create with

play18:43

it my free school Community is kind of

play18:45

the hub for AI automation agencies and

play18:47

you being able to identify Solutions and

play18:50

and really be around other people who

play18:51

are interested in making the most of

play18:52

this AI opportunity so that sounds

play18:54

interesting you can join the school in

play18:55

the description below excited to see

play18:57

what you guys can build with this if

play18:58

you're exced deed to get stuck into zapy

play18:59

essential and have got something out of

play19:01

this video please hit down below and

play19:02

leave a like on it um really supports me

play19:04

to continue making fre videos like this

play19:06

and of course if you want to stay up to

play19:07

date regarding future updates on zapia

play19:09

Central how to build gpts and other tips

play19:11

and tricks that I'm finding for you to

play19:13

make money and build businesses in the

play19:14

AI space be sure to subscribe while

play19:16

you're down there so that you don't miss

play19:17

any of my future videos if you're

play19:18

interested in learning how to start an

play19:19

AI business in 2024 I've just released a

play19:22

full guide an hour and a half long that

play19:24

I highly recommend you watching here but

play19:25

aside from that guys that's all for the

play19:26

video thank you so much for watching and

play19:28

I'll see you in the next one

Rate This

5.0 / 5 (0 votes)

Related Tags
AI自动化Zapia Central无需编码工作流GPT交互企业效率AI代理自动化平台数据集成创新技术
Do you need a summary in English?