[0:00] hey everyone Liam Watley here and today
[0:01] I have another Super valuable video for
[0:03] you all showing you how to build a
[0:05] $32,000 AI chatbot that's been built and
[0:08] sold successfully within our community
[0:10] so I'm going to be walking you through
[0:11] step by step I've basically done all of
[0:13] the work for you everything is templated
[0:15] what are you're going to need to do is
[0:16] click a couple buttons to integrate it
[0:17] all together and you don't have to write
[0:19] a single line of code so it's going to
[0:20] be super quick I'm going to keep it
[0:22] really brief but you guys can take this
[0:23] and sell it it's being sold by people
[0:25] successfully in our community which I'm
[0:26] going to go into in a second but you can
[0:28] take this and sell it to your own client
[0:30] and start marketing it as well so great
[0:31] thing to get started if you're just
[0:32] getting into the AI solution space and
[0:34] let's jump into it if you are new to the
[0:36] channel and don't know who I am my name
[0:37] is Liam otley I've been building and
[0:39] selling AI Solutions like this through
[0:40] my own AI automation agency Morningside
[0:43] AI for the past 18 to 20 months now I've
[0:45] been really screaming this from the
[0:46] rooftops that this is the way to get
[0:47] into the AI space for over a year and a
[0:49] half now so that everything I do on this
[0:51] channel is built around teaching you how
[0:53] to build and sell these Solutions and
[0:54] build a business around it so in this
[0:55] case I've done literally all of the hard
[0:57] work for you I've sat here for a few
[0:58] hours today rebuilding the system using
[1:00] no code software so that you can just
[1:02] click a couple buttons download the
[1:03] templates and connect it all which I'm
[1:05] just going to show you how to do in this
[1:06] video so there's no excuse now that uh I
[1:08] don't know how to build things I don't
[1:09] know what to sell I'm giving it to you
[1:11] right here so let's jump into it okay so
[1:12] what exactly are we going to be building
[1:14] well this is an AI market research
[1:16] assistant and this was sold by Abby aasu
[1:19] I think I'm pronouncing that correctly
[1:20] who was a member of our community and a
[1:22] couple of weeks ago I put out a form
[1:24] which is this form here um where people
[1:26] could submit their Solutions and let us
[1:27] know how successful they're being what
[1:29] they're selling and and providing the
[1:30] proof as well so Abby was one of those
[1:32] who applied on that form let us know
[1:33] about the solution is built and I'll put
[1:35] her proof of payment up on screen here
[1:36] now she said she got paid 75% upfront so
[1:39] this transaction is The Upfront amount I
[1:41] assume she's still in the building
[1:41] process for this now so Abby is the one
[1:43] who came up with this idea so if you
[1:44] want to get in touch with her you can
[1:45] work with your agency right here or
[1:47] click on this to be taken to the
[1:48] LinkedIn and if you want to be featured
[1:49] in future videos like this and get your
[1:51] name put in front of thousands and
[1:52] thousands of people then you can use
[1:54] this form here so you guys will get
[1:55] access to this figma on the scroll in a
[1:56] second so what we're going to be
[1:57] building is an AI market research
[1:59] assistant which is an agent that takes
[2:01] in questions from us the user and then
[2:03] sends those to three different agents
[2:05] which represent different personas or
[2:06] different users of a of a certain
[2:08] product or service so different
[2:10] businesses have different personas say
[2:12] we have the this type of user in my case
[2:14] myself s agentive we have ai newes some
[2:16] of you watching this might be completely
[2:18] new to building AI Solutions so that's
[2:19] one of the people that will use my
[2:21] software we have developers who want to
[2:22] use agentive to streamline the
[2:24] development process we also have
[2:25] business owners who have had maybe a
[2:27] developer work with them in the past and
[2:29] the Developers transferred onto their
[2:30] account and now the business owner is a
[2:32] user of agative so three different types
[2:34] of um people who or personas who use the
[2:36] software and this system basically
[2:38] allows us to send questions to these
[2:40] different personas and ask what they
[2:41] think about things like feature request
[2:43] or pricing and so inside each of the
[2:45] prompts and knowledge bases if you want
[2:47] uh you'll see in a second I've written
[2:48] these quite in-depth prompts on each of
[2:49] these to tell them who they are what
[2:51] they use the software for what they
[2:52] really want out of this what they want
[2:53] out of life um they are able to respond
[2:56] in line with how a as close enough as
[2:58] possible to a real person would reply so
[3:00] it allows for very rapid market research
[3:03] instead of having to get send out send
[3:05] out emails and ask for surveys and or
[3:07] like get people in a room and do little
[3:08] like test groups and ask them a bunch of
[3:10] questions which is expensive takes a lot
[3:11] of time you're dealing with people this
[3:13] is a far more rapid way of doing this
[3:14] and therefore can be very valuable to
[3:16] businesses if they're looking to get a
[3:17] soundboard on new features and and new
[3:19] things like that so um walking through
[3:21] the system here we're going to be
[3:22] creating three different agents on
[3:24] agenta my software don't wor it's
[3:25] completely free apart from your open AI
[3:27] usage cost so we don't charge anything
[3:29] it's just the open AI usage cost so
[3:30] you're getting the lowest cost possible
[3:32] so we're going to be creating three of
[3:33] these I've already built them already
[3:34] written the prompts so you guys are just
[3:36] going to be able to snag the templates
[3:37] for them and be able to mimic those
[3:38] prompts when you build your own versions
[3:40] so we have these three agents and then
[3:41] we have a master agent which is the one
[3:43] we're actually going to be talking to
[3:44] and this is what we really hand over to
[3:46] the client as the product um and we're
[3:48] going to be able to send messages to it
[3:50] it's going to take those messages send
[3:51] it to a relevance AI tool which I've
[3:53] built a template for you don't need to
[3:54] worry about um that's going to manage
[3:56] the sending of a request to these
[3:58] different agents and then when it all
[4:00] returns to the AI market research
[4:01] assistant it's then going to aggregate
[4:03] it and make a summary of all the
[4:04] different responses from these different
[4:06] personas and what the AI newb thought
[4:08] about the feature or the developer and
[4:10] the business owner and then give us a
[4:11] response saying hey this is what each of
[4:13] them thought about it maybe you could do
[4:15] this this this this so that's the system
[4:17] this is the step-by-step process we're
[4:19] going to go through to build it so all
[4:20] of the resources there's quite a lot cuz
[4:21] I've put quite a lot of work into this
[4:22] for you all it's all going to be in my
[4:24] free school Community it's going to be
[4:25] first link in the description it'll need
[4:27] you to sign up to agent of my free
[4:29] software you'll be be able to clone my
[4:30] agents all of these that are built into
[4:32] your account in just a couple clicks
[4:33] you'll then be able to sign up to
[4:34] relevance AI which they do have a free
[4:35] plan as well you'll be able to clone my
[4:37] relevance AI tool and connect it to the
[4:39] main agent this one down here and then
[4:41] you'll be able to test it and then
[4:43] you'll be done so there's barely any
[4:45] work to do here it's really just
[4:46] connecting the different pieces that
[4:47] I've already built for you so let's give
[4:48] you a little demo of what it looks like
[4:50] at the end if I say I paste this in I've
[4:52] got a little message I have a question
[4:54] about a client feedback mode feature
[4:56] that we've been thinking about for
[4:57] agentive so this is my agentive market
[4:59] res search system that's giving me
[5:01] feedback on from the Newbie from the
[5:02] developer and from the business owner I
[5:04] can now ask and say hey what do you
[5:06] think of this feature what do you think
[5:07] of this pricing model so I'm going to
[5:09] send this and see what it says back you
[5:11] can take a pause of the video and just
[5:12] read over what I've I've asked about
[5:14] there it's really interesting feature I
[5:15] would love to hear your thoughts on it
[5:16] CU I think it could be very handy for AI
[5:18] automation agency owners because I know
[5:19] from running my own agency how difficult
[5:21] it can be to send chatbots over for
[5:24] client feedback and then they send looms
[5:25] back and then they do all this anyway
[5:27] gets really messy that's why we're
[5:28] thinking of doing this client feedback
[5:29] mode with an agentive okay so we have
[5:31] three different perspectives here the
[5:32] developer says it is highly beneficial
[5:34] they appreciate the dedicated feedback
[5:36] page for client feature is really aimed
[5:37] at developers um they have the business
[5:39] owner perspective they are expressing
[5:41] reservations noting that the main
[5:42] concerns line usability and performance
[5:44] rather than developer client dynamic
[5:45] they prefer features that will Empower
[5:47] them to self-manage and optimize their
[5:48] agents without replying on a developer
[5:50] so business owners understandably more
[5:52] focused on getting maybe better costs or
[5:54] making it easier for them to self-manage
[5:56] not too worried about that develop a
[5:57] client relationship the newbies are
[5:59] pretty in as as well stressing the need
[6:00] for user friendliness especially for
[6:02] non-developers as that typically not as
[6:04] technical so you can see that we get
[6:05] three different perspectives and the
[6:07] prompting that I've done allows us to
[6:08] have quite like adversarial perspectives
[6:10] and what different people and different
[6:12] personas want out of the application so
[6:14] this kind of solution can be super
[6:15] helpful for getting rapid feedback from
[6:17] your different personas that are end
[6:18] users of your application without
[6:19] spending thousands of dollars on focus
[6:21] groups or without having to go through
[6:23] the hassle of actually organizing them
[6:24] so enough talking let's get into the
[6:26] build so we have a couple different
[6:27] agents we have firstly our Central agent
[6:29] the agentive market research assistant
[6:31] which is the one that we're going to be
[6:32] sending messages to then we have our
[6:33] three different personas the AI newbie
[6:35] Persona the business owner Persona and
[6:37] the developer Persona so in each of
[6:39] these I have written a detailed prompt
[6:41] and you guys can take this and apply it
[6:43] however you want to you can throw it
[6:44] into Claude or chat GPT and say hey take
[6:46] the same style and I want to do it for
[6:48] this so everything I've put in here
[6:49] about giving some information on the
[6:51] product providing some details on what
[6:53] they want out of it Etc and their
[6:54] background and some general context it's
[6:56] all really good prompting that you can
[6:57] take and apply to whatever you want so
[6:59] the first and you following along with
[7:00] this build is going to be going to the
[7:02] description joining the school Community
[7:03] finding the YouTube resources a little
[7:05] tab for YouTube resources you'll find
[7:07] this video you see the thumbnail on it
[7:09] and inside that will be all of the links
[7:10] you need to clone all of these agents so
[7:12] go now and clone all of these agents
[7:13] into your agentive account you will need
[7:15] to sign up if you haven't already so
[7:16] when you click the link it'll take you
[7:17] to the homepage you can sign up and
[7:19] register it's going to ask you for your
[7:20] open AI API key you need to put that in
[7:23] because everything we do on ative is
[7:25] built on top of the open AI assistance
[7:26] API so we're basically allowing you to
[7:28] build on top of that that API without
[7:30] having to code and do a whole bunch of
[7:31] unnecessary work so that's what we've
[7:33] built the gender for just to make it
[7:34] really easy to interact with that API so
[7:36] once you've gone and created your
[7:37] account added in your open AI API key
[7:40] and then cloned each of those agents
[7:41] into your account you'll have this this
[7:43] this and this and if you go back to our
[7:45] figma here you'll see these are the
[7:47] three agents the AI newbie Persona 1
[7:49] Persona 2 Persona 3 and the AI market
[7:51] research assistant now we just need to
[7:53] fill in the connecting part which is
[7:55] this relevance AI tool so if we go back
[7:57] to here there we another link on the
[8:00] school Community to get this template
[8:01] for an AI Persona query so what this
[8:04] takes in is the question that you want
[8:05] to send to the agent and then it also
[8:07] takes in the assistant ID so on agentive
[8:09] each of your assistants has an assistant
[8:10] ID which can be found by going here to
[8:13] make sure you've already published the
[8:14] agent and then you can click deploy you
[8:16] can go to custom code and scroll down a
[8:18] bit and you'll see the assistant ID is
[8:20] available here so each time we send a
[8:21] message to our executive agent here it's
[8:23] going to be taken the question we
[8:25] provided and then the assistant IDs for
[8:27] each of these agents which we're going
[8:28] to include in the prompt I'll show that
[8:29] in a second and it's going to query each
[8:31] of these agents through our relevance AI
[8:33] tool cuz you can see here it takes in
[8:35] the query and it takes in the assistant
[8:36] ID and if you go to the prompt on the
[8:38] market research assistant if we go to
[8:40] edit here and you scroll down you'll see
[8:43] that I have all of the keys in here so
[8:45] that's the next step for you now is to
[8:47] come into your agent of market research
[8:48] assistant so the main one we're going to
[8:50] be sending these messages to go to each
[8:52] of your agents we can start with the AI
[8:54] newbie Persona you can go deploy custom
[8:56] code and then copy this assistant ID
[9:00] I'm going to go back here scroll down
[9:02] and then AI newb replace this paste it
[9:05] in there so you need to do this and
[9:07] repeat the process with each of them
[9:08] make sure it's published click deploy go
[9:10] to custom code take the assistant ID
[9:13] take it back over to the market research
[9:14] assistant and make sure you've pasted in
[9:16] all of your assistant IDs to replace my
[9:17] old BS once you've got the prompt all
[9:19] set up ready to go you'll be able to go
[9:20] onto the school Community again and find
[9:22] the link to this exact relevance AI
[9:23] template there'll be a button up in the
[9:25] top right that will say clone into your
[9:27] account so you can edit it you'll be
[9:28] able to clone it onto your relevance AI
[9:30] account and then you'll see a screen
[9:31] that looks a little bit like this now
[9:33] the final step of getting this relevance
[9:34] tool set up I've done most of the hard
[9:36] work for you is to open up each of these
[9:37] API blocks and then there's going to be
[9:39] an API key here and that's your agentive
[9:41] API key so to get that we can come back
[9:43] to any of our agents in this case just
[9:46] on here and if we go to deploy custom
[9:48] code again we can see that our API key
[9:50] for agentive is here copy that head back
[9:53] to your relevant AI tool and come to the
[9:55] API key and paste it in here make sure
[9:58] you delete that and replace that with
[9:59] yours and the same for the next block
[10:02] down here which is going to be this one
[10:04] here can delete that and paste it in and
[10:05] just a little explainer if you're not
[10:06] sure what's going on here when we send
[10:08] in the query or the question that we
[10:09] want to ask to this panel of researchers
[10:12] we also send the assistant ID which our
[10:14] our our main agent is going to be
[10:15] sending for us from the prompt it's
[10:17] going to be calling each of your agent
[10:18] of assistant first it's creating a new
[10:20] session then we get the session ID out
[10:22] of that then we pass it into our chat
[10:23] endpoint which is going to send off the
[10:25] final payload to actually generate the
[10:27] response so don't want to get too Stuck
[10:28] In The Weeds there but once you've
[10:29] cloned this template you've gone through
[10:31] and added in your API key from this from
[10:33] the deploy tab here in the custom code
[10:35] section and you've added that to both
[10:36] the API blocks you're then ready to give
[10:38] this a test if you'd like you can come
[10:39] back to any of your agents and we can
[10:41] grab the assistant ID from one of them
[10:42] and then we can go back to relevance we
[10:44] can pass an a query
[10:46] um thoughts on a
[10:50] $100 plan with unlimited
[10:55] usage put that in there and if we go run
[10:59] off up in the top right at the bottom we
[11:02] should see the final thoughts of in this
[11:04] case the AI newbie agent and there you
[11:06] go you can see we're getting good
[11:07] responses from our tool the last step
[11:09] here to integrate this AI Persona query
[11:12] tool with your main agent is to make
[11:14] sure you've saved it in the top right
[11:15] here then we can go back we can click
[11:18] the new button up here and then we'll
[11:20] see this custom actions GPT button pop
[11:22] up here so this is why relevance AI if
[11:23] any of you guys are just getting started
[11:25] in the space highly recommend relevance
[11:26] AI is an excellent place to build tools
[11:28] because anything that you build in here
[11:30] and they've got a bunch of templates as
[11:31] well this is not a sponsored video by
[11:33] the way um you can then click this
[11:34] custom actions GPT button which allows
[11:36] you to put it onto gpts or in our case
[11:38] onto agentive so we can put it in
[11:39] different places as well so to do this
[11:42] we can just scroll down and click on our
[11:43] AI Persona query you can select multiple
[11:46] which is really handy they'll bundle it
[11:47] all up into one schema then you want to
[11:49] select the custom or here um I'm just
[11:51] going to generate the schema first we
[11:53] click this now we get our schema this is
[11:55] needed to connect the relevance tool to
[11:58] the agent so we can copy this
[12:00] we can come back to our agent of market
[12:02] research assist our main one we're going
[12:03] to be chatting to and click on the tools
[12:05] tab I already have a tool built in with
[12:07] this template but you can just click
[12:08] here and delete it CU you'll need to
[12:10] update it to your own we can click
[12:11] upload tool here paste in that schema
[12:14] it's going to add this AI Persona query
[12:17] tool and then to make sure we're
[12:19] authenticated to actually call relevance
[12:20] Ai and be approved to use it we need to
[12:22] click the authentication button we can
[12:23] click API key here we're going to go
[12:25] custom go back to relevance AI scroll up
[12:29] and generate our own API key here copy
[12:32] the API key come back paste it in the
[12:35] API key section and we're going to write
[12:38] authorization with an American spelling
[12:40] with the Zed in it
[12:42] authorization then we can click update
[12:44] authentication create tools and now we
[12:46] have connected our relevance AI tool
[12:48] which can take any assistant ID into it
[12:50] and generate a response and we've got
[12:52] all of the updated assistant IDs in here
[12:54] if take a moment to read through the
[12:55] prompt I've basically instructed this
[12:57] agent to hit that same tool the
[12:59] relevance thing that we've just set up
[13:00] hit that thing three different times
[13:02] using the assistant IDs in here so every
[13:05] time you get a message from the user
[13:06] make sure that you are taking one of
[13:07] these assistant IDs taking the user
[13:09] query that we just sent to it sending it
[13:11] to the tool getting the response and
[13:12] then doing that two more times with the
[13:14] other two tools in here you could set it
[13:16] up on three different tools but for
[13:17] Simplicity sake and if you guys watching
[13:19] this following along it's actually
[13:20] easier for you to just use one tool and
[13:22] and ping the thing three times so now
[13:24] that we've got all that set up we can
[13:26] click publish in this case
[13:29] and we can start sending it messages so
[13:31] um what do you think about a $100 per
[13:37] month um
[13:39] analytics plan which gives much deeper
[13:45] insights on uh your agent usage when
[13:49] processing a lot of volume so now it's
[13:52] going to take that query it's going to
[13:53] go to our relevance AI tool which you
[13:55] know by now takes in a query and an
[13:57] assistant ID and so what this is going
[13:58] to be doing is taking that query taking
[14:00] these assistant IDs that we' provided
[14:02] here sending it off and getting a
[14:03] response from each of our different
[14:04] personas bringing it back and the magic
[14:06] of plugging it all into one agent here
[14:08] on on agentive is that the llm behind
[14:11] the entire system is then going to take
[14:12] each of those responses and then based
[14:14] on our prompting here rewrite that all
[14:16] into something that we want and that's
[14:18] much more processable so instead of
[14:19] having these big blocks of text it's
[14:21] going to write a nice concise summary
[14:23] and here we go here we have our
[14:24] responses I can probably Zoom this up a
[14:25] little bit developers see the potential
[14:27] and benefits in the analytics plan the
[14:28] business acknowledges that deeper
[14:30] insights could be valuable particularly
[14:31] for operations involving significant
[14:33] volume however they express concerns
[14:34] about the cost effectiveness of the plan
[14:36] and desire a clear demonstration of its
[14:38] potential return on investment so the
[14:39] business owner of course very worried
[14:41] you can see in the prompts of each of
[14:42] those and the templates I've provided
[14:43] how I've prompted them into business
[14:45] owner being worried about cost and
[14:46] managing the thing themselves the AI
[14:48] newb is very interested in anything that
[14:49] can allow them to prototype and test but
[14:51] they're not really interested in the
[14:52] later stage features just yet they're
[14:53] more selfish and want uh the
[14:55] userfriendly beginner friendly stuff to
[14:57] be able to build and prototype quickly
[14:59] so you can see here the AI newb is
[15:00] interested in the idea of the analytics
[15:02] plan but feels it may be premature for
[15:04] them at this stage they recommend a
[15:05] gradual introduction to analytics
[15:07] features such as teer pricing or a free
[15:09] tier to facilitate their Learning
[15:11] Without immediate costs so this is just
[15:13] an example showing it on my own business
[15:14] it's I built it because it's going to be
[15:16] helpful for me I can now send things to
[15:17] this um and my team can use it as well I
[15:19] can provide it to the dev team and
[15:20] personally as a business owner using
[15:22] this for myself this is extremely
[15:23] helpful and I can only imagine if I was
[15:25] actually having to pay thousands and
[15:27] thousands of dollars a month to get this
[15:29] kind of insights from focus groups then
[15:31] I'd be definitely willing to spend a few
[15:33] thousand on a system like this for
[15:35] internal use and it's so much faster
[15:36] than having to use humans for it just
[15:38] quickly I forgot in the original
[15:39] recording to show you how to get that
[15:40] final result that I showed you which is
[15:42] a a full page way to chat with your
[15:43] agent so you can come up to here and
[15:45] click on the share button you can click
[15:46] the make public then you can click this
[15:49] I've copied it now I can open up a new
[15:51] tab and here I have my full page genen
[15:54] of market research assistant where I can
[15:56] say hello and send messages here so this
[15:58] link that you get
[15:59] um via this option here you can then
[16:01] send to your clients um for them to be
[16:03] able to chat with it and use it as they
[16:05] wish you can also send the template via
[16:07] email and you can generate a public link
[16:08] to share your templates around but this
[16:09] is the main one in order to get this
[16:11] full page chat wiget just going back to
[16:13] the figma here I want to say again thank
[16:14] you to Abby for sharing this with us if
[16:16] you want to be featured in more videos
[16:17] like this and some of the other list
[16:18] videos from sharing AI Solutions you get
[16:21] a massive shout out on my channel more
[16:22] traffic to your agencies more clients
[16:24] more people coming to work for you and
[16:25] in general if you're looking to
[16:26] replicate this build and do it yourself
[16:28] um this fig board will be on the free
[16:29] school Community as well so everything
[16:31] you need to do this will be on there and
[16:32] you can see the step-by-step process
[16:34] here if you get lost at all so I know
[16:36] some of you might be saying oh $32,000
[16:37] is's no way I could sell it for that
[16:39] that to some degree is a skill issue
[16:41] aby's found someone who's willing to pay
[16:42] that and in your case if someone's
[16:43] willing to pay $322,000 I bet you
[16:46] they're going to be willing to pay three
[16:47] or four so even if you don't think this
[16:49] exact product that I've shown you here
[16:50] is worth $32,000 you might need to take
[16:52] another step further and take it from a
[16:53] genive and put it onto a custom front
[16:55] end that seems like more white labeled
[16:57] and looks better for the client or
[16:58] connect it to say slack or something you
[17:00] you might need to add more value on to
[17:01] it but there's definitely people and
[17:02] businesses out there as Abby has proved
[17:04] who are willing to pay for these things
[17:06] and customer research and market
[17:07] research like this is a massive function
[17:09] for businesses especially as they get
[17:11] larger so take this and do with it what
[17:13] you will I can't do much more than spoon
[17:15] feed this stuff to you without no code
[17:17] no nothing tutorial everything templated
[17:20] please guys I'm I'm doing everything I
[17:22] can here to give you a leg up and get
[17:24] into this because it has changed my life
[17:25] and helped me transition into a career
[17:26] where I love what I'm doing every day
[17:28] and working with people that I really
[17:29] love being around so you guys know I I
[17:31] don't want to hop on for too long but
[17:33] this has been a tutorial showing you how
[17:34] to use my software agentive to create a
[17:35] market research assistant for my own
[17:37] soft agentive all of the links available
[17:39] in the description and there's going to
[17:40] be a tutorial up here showing you how to
[17:41] take exactly what you've just built here
[17:43] and put it onto WhatsApp using our no
[17:45] code WhatsApp integration on agentive
[17:46] and just a couple Clicks in a few
[17:48] minutes so check that out if you want to
[17:49] add on to this and create more value for
[17:51] your clients by putting it on WhatsApp
[17:53] but aside from that that's been all for
[17:54] the video guys I'm running out of
[17:55] daylight here but I will see you in the
[17:56] next one