Leveraging AI

185 | Boost Sales With an AI Agent That Researches, Reports & Preps Sales Calls with Ashleigh Stearn

• Isar Meitis • Season 1 • Episode 185

Share episode

👉 Fill out the listener survey - https://services.multiplai.ai/lai-survey
👉 Learn more about the AI Business Transformation Course starting May 12 — spots are limited - http://multiplai.ai/ai-course/  

AI agents that do the heavy lifting in the sales process sound amazing “… but how do I actually do this?”
If you asked yourself this question - this session is for you!

Join us live as Ashleigh Stearn, a Relevance AI-certified expert and founder of a growing AI mentorship community, walks us through a real AI agent she built that transforms the sales process. From CRM scraping to prospect research to pre-call reports and automatic updates — this single AI agent does the heavy lifting so your team doesn’t have to.

Ashleigh has logged over 900 hours mastering Relevance AI (a No-Code agent development platform), all without a technical background. Her superpower? Making complex workflows simple. She’s helped countless professionals break through the AI learning curve — by translating overwhelming systems into digestible, actionable steps.

In this session, you’ll get a behind-the-scenes look at a working AI-powered sales workflow — with clear explanations of how it’s built, what connects to what, and where the biggest roadblocks (and breakthroughs) happen. Expect clarity, honesty, and a few “wait, AI can do that?!” moments.

About Leveraging AI

If you’ve enjoyed or benefited from some of the insights of this episode, leave us a five-star review on your favorite podcast platform, and let us know what you learned, found helpful, or liked most about this show!

Isar Meitis:

Hello and welcome to another live episode of the Leveraging AI Podcast, the podcast that shares practical, ethical ways to grow your business Improve efficiency and advance your career. This is Isar Meitis, your host, and we have a really exciting show for you today. We've said many times since 24, and definitely in the first few months. So this year that 2025 is gonna be remembered as the year of agents. And this has turned to be completely true. Everybody wants agents, everybody wants to deploy agents. Everybody wants agents a part of their workforce. The problem is the vast majority of business people don't know what agents are and even less know how to actually build them and what it actually means. What are the steps you need to take, which tools you can use, and so on. So what we're going to do on today's show. Is demystify it and actually show you how you can build your own agents. But we're going to dive even deeper into that. We're gonna show you a specific use case on how you can build a team of agents that can support your sales team, that can do research, that can grab information, that can update your CRM, that can do a lot of things that are going to help your salespeople do their sales job better. Now, the really cool thing about all of this is it's done without writing any code and in tools that literally anybody can use. And even better is that our guest today, Ash Stern, is a certified expert and partner of relevance.ai, which is one of the platforms that a lot of people use to build these agents, and she has been developing and deploying agents for multiple companies. In multiple industries in the past year. So she has been doing this for a living in the trenches, using relevance for a while, and she's an amazing expert and a great teacher that can explain exactly how she does it. So I'm personally really excited and I know a lot of you are too. So Ash, welcome to leveraging ai.

Ash Stearn:

Thank you. Thanks so much for having me. Very excited.

Isar Meitis:

I am as excited as you. I think I'm actually more excited than you because you know exactly what you're gonna show. And I'm like, I want to see it. I wanna see, and I'm sure a lot of people are the same way. And so, great stuff for all of you who are joining us right now on Zoom and or on LinkedIn Live, first of all, welcome. Thank you for spending this time with us. We're very excited to have you, go introduce yourself, say where you're from, say kinda like where you're in your AI journey, what do you know about agents? So we know where you are and we can relate to that as we do our show. If you're not with us live, the question is why aren't you with us live? We do this every Thursday at noon pm Eastern with amazing people like Ash. we're gonna dive with each and every one of them into a practical AI use case. And if you're here a, you can network with amazing people who are business people who care about ai. And also you can ask questions, which you cannot do if you're doing this while listening after the fact. So that's. One item. The second thing that I wanna say before we dive into this topic is that we. Have started the registration for the next cohort of the AI Business Transformation course. It's the course that we have been teaching for two years now. It's crazy that it's been two years, but since April of 2023, I've been teaching this course myself. I. Online on Zoom with cohorts of people. And so this cohort starts on May 12th. There has been hundreds of business people that literally transform their businesses and or their careers with the knowledge they acquired in the course. And so the course obviously updates every single time we run it. We run it mostly privately for different companies and organizations. I'm actually teaching two private courses in parallel to this public course, but the public course starts on May 12th. And we do the public courses only once a quarter. So if you haven't done any proper training for yourself, you owe it to yourself to do proper AI training so you can really transform everything that you're doing across multiple aspects of the business. there's gonna be a link in the show notes, and then we are gonna drop the link also in the chat right now. And if you use promo code leveraging AI 100, so the name of the podcast, 100, you will get$100 off just for listening to this podcast. How cool is that? So that's it about the course. And now let's dive into agents, how they're built, how they're coordinated with themselves, and how they actually are working. Ash, it's all yours.

Ash Stearn:

Awesome. All right, cool. So what I'll do is I'm gonna jump straight into the platform and actually show you, but kind of explain everything as we're going. and I think that's the best way to do it. I do like presentations and slides, but there's nothing better than actually putting two and two together by actually going into the platform itself. So I'm just gonna, share my screen and we can get to that. Okay. So what we've got here is we're gonna essentially start off with the team that we've got Now, our team currently is sitting under our manager, which is manager Mitch. Now when we come down into this is the agent dashboard, essentially, this is what you're gonna see as soon as you start creating your very own agent. now before we kind of get into, I guess, the details of how it's all set up, I'll essentially show you, what it looks like. All right. So just some context here. this was set up, and was connected to my Calendly link. So what would happen is that somebody would book with me through Calendly, they'd fill out a form, which included obviously these details that we've got up here. And then this would actually be triggered and sent directly to my agent to start the entire research process. So when we click, so this is the information, we click down and we can actually see our manager, which is who we're currently in at the moment. I'll pause you just

Isar Meitis:

for one second. For those of you who are just listening and not watching this, two things. One, there's a YouTube channel that you can find very easily by clicking in the YouTube channel link in your show notes. So if you wanna watch this, and you can do that. But if you're driving or walking your dog, or doing the dishes or whatever it is that you're doing, listening to podcasts or yoga or I don't know what you're doing, we'll explain everything that's on the screen. So in this particular case, we're looking at a screen in, Relevance and on top there's information that came in that triggered the agents to start working. So it's the email, first name, last name, meeting type. Again, all the information that basically come Oh, time zone, the information that comes in from a meeting that you set up on Calendly.

Ash Stearn:

Right. Right. Exactly. So, so with that information that we've got, our agent, you can actually see the process of our manager agent delegating this information to our sub-agents. Now we've got two sub-agents. Okay. You can actually see here that it's delegated to our lead researcher and then delegated to our CRM update agent. Okay. So essentially what our manager is doing is that it's communicating, it's being programmed in a way to communicate and delegate tasks to our agents. Our agents do what they need to do. So obviously our lead researcher is re doing all sorts of, research based on our lead. And then we've got our update as well into our CRM. So this is the information that we get from all of that research. So information includes, the email, job title, LinkedIn profile. We've got company information, such as the website, LinkedIn page, a company summary. And then we've also got a pre-sales call report. Now, I really wanted a pre-sales call report written for me so that I was very well prepared before going into the discovery calls. Now there's nothing more off putting than going into a sales call and just not knowing who you are talking to and just how you can really serve them. So our pre-sales call report, there are two ways that you could do this. You could do it within relevance ai, but at this time, when I did, this particular setup, I actually had it connected to Make, now, if you haven't heard of make is an automation platform. It's a very popular one. So it's a visual builder and it allows you to visually build out workflows. Essentially make is you are selecting a whole bunch of different nodes with make, and they are connected, right? So you're connecting all of these different components in make that make up an automation. Okay? Now, when creating agents, you need to understand that automation is a part of that. and so I set up a little automation. That's added my presale score report to a Google document. Now, the wonderful thing about relevance AI is that they've now, added in a whole load of integrations that you can now add, to your agents directly. And one of those is Google documents. so essentially you're not gonna need any of these external integrations with like other platforms, like Make and Zapier'cause, like one that's annoying. Two, it's technical as well if you're not familiar with those platforms. but just to show you how the report was actually structured, this is what it looks like now. This is actually being done within Relevance ai. that will actually give me the link to, to my Google Doc without going through make and stuff. So we've got top expert themes. We've got a pre-sales, strategy, so rapport, building elements, professional achievements. Then we've got pain points and solution alignment. Value proposition and then shared content and priorities. Now, I like this because I like to see how active my lead is on LinkedIn, and if they're active on LinkedIn, obviously if they've got recent content, it really paints a picture of like, what are they interested in and, in terms of business and what they're doing. So I like that. so essentially this is what our report, looks like. now if we go back into quick question.

Isar Meitis:

Yes. on the pain points side. Yeah. how does it know their pain points? Does it do research on its own? Is it based on what they wrote in the form? Is it a combination of these two? Like, what is the process to get that.

Ash Stearn:

how the pain points are done is that you actually have within your particular agent who's creating the presales core report. In this case, it's my lead researcher. They actually have my information. So my information is then combined with their information and just the research that it's done based on that lead so that it can perhaps find some pain points where my services and what I offer is going to be of value to them. I have found that sometimes people will actually post content on their pain points, like what are they struggling with in their business and stuff like that. So that is another indicator of pain points as well. but our lead researcher does all of that for us, which is really cool.

Isar Meitis:

And we're gonna dive into that afterwards Right, exactly. To see what it's doing under the hood.

Ash Stearn:

Yeah, exactly. So we'll get into more detail about that. Right. So that is essentially, our team, a little mini team of agents. Now, it's not a big team. Some of these teams can get quite complex. but one of the main things we need to keep in mind here, and what we'll do is we'll actually go into the lead researcher now, just so I can really show you exactly how it works. Now, within relevance ai, you need to try and understand what's your workaround, how are you going to actually build agents, what's your interpretation of agents? how do you wanna approach building agents? Relevance ai, as you can see here, is really based on prompt engineering. Now, this is one of the things that people need to understand is that if you're going to build agents, whether it's with relevance, ai or another platform, prompt engineering is a part of that. Your agents are built on, LLM models. So prompt engineering. Just to set an expectation, here is a skill that you'll need to have when building any kind of agent. If you're using LLM models, you need to know prompt engineering. so relevance ai, the agents here that we build, the brains of them are the core instructions. So the core instructions are essentially, one big prompt. And in this, particular case, what we use is a prompting technique called chain of thought prompting. So it is a very common way to prompt LLM models these days. What it allows us to do is that it allows us to really help our agent understand the process it needs to go through in order to get the best output. So as we're looking at our lead researcher agent, we can go through this prompt and see how I've actually programmed it to work. All right? So we've got a role, so we're going to always give it a role. we then have an objective, okay? We then have context as well. And now what's different with an agent prompt in relevance AI is that you're gonna have an SOP. Now usually if you're just using this prompt with an LLM model like chat, GBT or clawed, or whatever else it might be, you're not usually gonna use an SOP. You might use instructions, okay? And only just instructions. but in this case, because we're building an AI agent, an SOP is good to have in there. So we have that. And within this SOP we're really getting quite granular on the different, steps it needs to be doing, the, process it needs to go through. So we've got our tools tagged in there as well, which you can add by just doing a backslash and it will actually add in the tool for you. And then we've got just some very brief overall instructions of just the overall tasks it needs to do for us. Then of course, we list, so I wanna pause you just

Isar Meitis:

for one second. Yeah. So, how did you come up with this, right? Because you've been doing this for months and you have a lot of experience. If I'm just getting started. And I wanna build one of those, how do I get started from scratch? First of all, do they have templates scratch, or do I ask Chachi PT for recommendations? Like what's the best process if somebody just getting started, how to figure out, because we just named the topics, right? We said, okay, there's objectives and there's context and there's SOP and there's a general instructions and then, but what actually goes in them? And we can obviously read this to everybody, but that's not the point if, because people wanna build something else, what you built, right? So how do you get started? if you don't know what you're doing,

Ash Stearn:

okay. If you dunno what you're doing, then I would actually get started with what is a process you're currently doing for your business that you don't wanna do. so if you have a process currently that you don't want to do and you want to, streamline with AI agents, then that is the first starting point. That's where I started and that's why I built this lead researcher. I didn't wanna waste time researching leads. I wanted to actually take more time in preparing for my meeting and doing other things like building agents. So what I actually recommend to people is going through the current processes that they're doing and seeing what can actually be, automated straight away with AI agents. and then go from there if you've got the process in detail, right? Even better, because then you're really getting quite granular on what is the process, what needs to be automated, and how you're gonna build towards that. Now with your agents, you've got building your agent for one particular process. In this case, what we're looking at is we are researching leads, but within the lead researcher you've got. Tools. Now your tools are, what are your automations? And now this is what I mean by understanding that. building agents, you have to develop multiple skills. so if we take a look at, let's say our pre-sales core report tool, which is one that is responsible for actually writing out report, you'll actually see that this is an automation and it's very similar to what you'd see in another automation platform like Make or Zapier. and essentially our tools are the skills for our agents to leverage and use to perform their tasks. But we have to build these. So when you think of a lead researcher, you think that's their role. So you think of a role in a human sense or a human department sense. Let's say for example, what I was doing was content production or Social media management. Within that social media management role, I had to. Write content. then I'm gonna build an agent that's gonna write content. So it's very similar in this case where lead research, that was one entire role. I didn't wanna do that. So I'm going to, automate that with an agent. And then what are the mini tasks within that role? So we have research, we have generating the presales core report. and then we also have verifying the leads email as well. those little mini subtasks, you then have to create tools to do them. so

Isar Meitis:

I want to pause. You wanna pause you for two things. one yeah. Is I want to give some idea for people'cause you said something. That is brilliant and we didn't dive into that. And I think it's important to say, start with by mapping your existing processes and people are like, yes. Oh my God, that's a great idea. But like, how do I do that? And so, the way I work with my clients when we map processes is I literally ask people to narrate what they're doing as they're doing it. So open your screen. Open the tools you're using. Hit record on whatever record platform that you want to use. You can use, the usual suspects. but you can use Zoom. Like you can have a meeting in Zoom on your own and just run that, and just record the screen and explain what you're doing as you're doing it. Do it three times. If there's multiple people doing it, ask each and every one of them to do it three times. Upload all of that into your favorite LLM Could be Gemini Clause, Chachi pt. I don't care. And say, Hey, I want you to create an SOP of this process. Here are transcriptions of recordings of Six different occasions doing it.'cause then. People do it a little differently, different times, and it will create an amazing SOP. Then you wanna ask it, to ask you questions to figure out if there's any gaps. It's gonna ask you questions, you're gonna answer them, and you have your SOP. That could be a very solid starting point for your agent. So that was, I wanted to add, what I want to ask about these tools is I assume like a section here that is creating these tools, right? And so can you explain what does that mean to create a tool? show us a little bit more detail on that? Absolutely.

Ash Stearn:

Yeah. So when we come into our dashboard, this is relevance ai. This is your profile. This is how it's all gonna look when you're in here. You're gonna come to your tool section, and now you're gonna see a whole bunch of tools, that I've actually created. all of these tools are where they, all live. now you can build them from scratch, but you can also use templates. relevance ai, when they first started, they actually only had six tool templates. now they've got an entire library of both agent templates and tool templates. these are great to start using, especially as you are learning and getting started. And in fact, how I learned, quick or quicker than, trying to rely on documentation. There's documentation back then and content. To learn how to do this was, not very good'cause it wasn't a very well known platform at the time. But I had six tool templates and all I did was I clicked on one, let's say for example, perform Google search. So I clicked on it. I then have to, clone the tool to be able to see what's going on in the back end of it so I can understand how they actually did it. And then I would actually create a tool completely from zero from scratch. And I just copied, so I just copied these steps. in our tools, we've got our inputs, and then we've got these different steps that we also have to add in there as well. Okay. Now this one's a very, a small tool, but it's, so it's basically

Isar Meitis:

like a smart flow chart, right? You start with an input and then what if, or what steps you need to take. And that becomes a tool that then the cool thing is, I guess they're all reusable because any of the agent can use any of the tools that you created, right? So if you create them generic, then you can use them across other tools as well. So in this particular case, a Google research, while you wanna do research across probably multiple different kinds of agents.

Ash Stearn:

Exactly. and obviously a Google search, tool can be used across. All sorts of domains as well, so it's not, tied to one domain either. so they can be used quite regularly by various different agents as well. And what, something is really important here to know is that how do our agents actually use these tools, right? So we know that they get given information, okay, but they actually dynamically give or use that information that has been given to it, and it will actually fill out the inputs. By themselves. We don't have to put in any information or anything like that into this area. Our agent actually does all of this for us in order to use the tools, effectively. so they dynamically add the information to our tools for us. and it doesn't matter what type of input it is, it could be a text input. So in this case it's gonna be a Google search query, but then we've got various other different types of inputs. So numbers, we've got files. We have an options menu. So yes, no, and our agent will actually fill all of these different types of inputs out by itself. what I would generally do with this was that, okay, so I cloned it. I then go back into my dashboard. I'd go back to my tools here. And then I would create a new tool. And what this helped for me when I was learning how to build agents was that it allowed me to understand what needed to go where. So I would go back into the tool that I've cloned, I'd understand, okay, well this is the title, right? And I'd simply just copy and paste When I needed to add a step, which was Google Step, I'd come into here, I'd add a step and I'd click Google Search. Okay. now there are a lot of little nuances to it, including variables. So it does get quite technical. And this is another thing I want to set an expectation for as well. It's technical in nature. it is something you're gonna have to learn and the learning curve can be quite steep. so in saying that, can non-technical people, build AI agents? Absolutely. You are looking at one. I was writing blogs before I was doing this, so yes. But don't underestimate the learning curve, because if you consider yourself a non-technical person now, well by the end of all of this you'll be considered a technical expert. So, just wanna set an expectation there.'cause it's not, something you're going to probably learn in two weeks. depending on your learning style.

Isar Meitis:

I wanna add two things to what you said because I think it's really important, two aspects of what you said that I think that are critical. One, it's perfectly fine starting by copying. Others, and others could be templates from the platform. Absolutely. Or people who share how they do their thing. And a lot of people do that. That's what we're doing right now. Right. Ash is literally showing you the agent, like you can watch the video copy word to word what you wrote, and you have an agent, a starting point. So it's perfectly fine to start by copying others because it allows you to understand the nuances and the processes and the little things that if you just try it on your own, will take you a lot longer. Yeah. And two is the ROI is still there. Yes. We're like, oh my God, now I gotta invest a month to learn this. Yeah, but this is gonna save you a month, every month. Then it's worth investing the first month in learning the platform. On the bigger, broader sense, from a company perspective, from a team perspective, you need one person who needs to know how to do this. You don't need everybody on your team. You don't need everybody in the company to know how to do this. And there's two ways to do that. One, you can hire Ash, right? you don't have to have that person in house. You could be an external person that knows how to build that, that can build it for you. two, you can get one of your team to be certified or train on how to use, either relevance or other platforms to build agents. And then that person can be the go-to person in the company, to build a platform. That obviously means they need time and resources and other stuff, but I'm putting that aside for a second. the learning curve is worth it. Because the yes variety of agents you can build can help across almost everything in your company.

Ash Stearn:

Yeah, and I will just add something here as well, with the speed of how everything is advancing at the moment. These are only gonna become quicker, more reliable. They're gonna get better. and they're eventually gonna be cheaper, hopefully in the future as well. so it is something that is constantly improving. And I've seen relevance AI where when they first started to where they are now, and just the amazing leaps that have come in the, incredible updates that they've done recently. They're making it, the platform is specifically targeting industry experts. they're not targeting technical experts. It is a platform that they're trying to build into something that anyone, any industry expert can come in, say they want an agent and have the agent, built. they've actually got a new feature that was just, released not too long ago where you can actually create, you can invent a tool. You can also invent an agent just by putting in a prompt.

Isar Meitis:

and you will write all the details and all the steps and all the other stuff.

Ash Stearn:

it's gonna write the core instructions. Now the core instructions and the prompting that goes into that is actually a very creative side to agents when you're building them with ring relevance, ai. And I always say to people and recommend, experiment with it. just because my way has worked for me doesn't mean that another way of writing the core instructions is gonna work. or it's not gonna work. it, there's no right or wrong here. And I think that just goes to show that the platform is flexible and the agents aren't intelligent. They will get it. As long as the instructions are clear and they're clearly articulated, they will get it.

Isar Meitis:

let's do two things because I'm very curious. I'll start with a question and then that will probably guide my next step. When the agent is actually running, can you see what it's doing? Like, can you follow the steps as it's doing the steps?

Ash Stearn:

what we can see happening once, a task has been triggered, we can see what's being performed in the background. So these are the tools that this agent is using. so it's used, obviously verifying the leads email first because in this case I was getting both, leads coming in that were filling out with a business email as well as a personal email. Now, that was really core in how we extracted the domain to understand the company that they worked for. So I had to verify and then it went through then and research the lead if the email was a business email. And then when we actually click on the tool that's being, used in this case. This one here, we can actually see the inputs that have gone in. So our agent has done that for us. We can actually see then the outputs that have been generated from this tool. We can't see what's going. So again, for those

Isar Meitis:

of you who can see, it got an email address. It checked if it's a business domain, and then it went and did the research on the company by extracting the domain name. So he came back and said, this is a, financial services company, the company URL on LinkedIn, A summary about the company, and stuff like that. Probably top people in the company and things like that, right? So, decision maker yes or no about this particular person. So it pulls all that information because it is a company domain and then it knows how to find information about it, and that's very cool.

Ash Stearn:

Yeah. Yeah. And this was another thing that I wanna mention as well, that sometimes you don't think of the challenges or the obstacles that are gonna come up until you actually start testing your agent. I had to think of a workaround. I have to now understand that what if someone uses their personal email? Obviously if they're extracting the domain of Gmail or outlook, it's not going to do the proper research because that's not the company that the person works for usually. so, the tool that I've done in that case is gonna be structured slightly more different to what how this one is structured.'cause there needs to be a workaround for how it's going to figure out. Where the lead, who the lead works for. and that's something I didn't really put two and two together until I actually started testing the agent out, and then had to accompany for that. So yeah, it's just interesting how it all falls into place and you gotta change things.

Isar Meitis:

so the first step, it gets an email, it researches the email, it gets this kind of information. What does it do next? I just want to go step by step so people understand what the agent actually does on its own.

Ash Stearn:

Sure. What we can do, we'll start off with the verify email first. So let's open this one up and then we'll get into the research tool. So this is all coming from the lead researcher, by the way. these are all the tools that are have been added to this lead researcher agent. So the first step it's always going to do within that SOP and the core instructions I've laid out, it's gonna use our verify lead email tool. This one is just a very simple tool, is an LLM step in here that we've added. We just need to clarify, is this a personal email, or is this a business email? Okay. So in this case, we've added just a simple LLM step and give it instructions. So it's not a very big prompt, it's just a pretty medium sized prompt instructions of how it needs to figure that out. Some examples as well. Examples are always very handy to have in your prompts. And then just some, notes don't give me pretext or context and only, output the correct classification. And that's all we need to do for that one. Once it's done that it's gonna output the, Whether it's business, the Yeah, whether personal, whether it's business or personal.

Isar Meitis:

Yeah.

Ash Stearn:

Right. Exactly. And based on that answer, our agent is then going to use the next tool. So in this case it is a business email. So it's now going to use the research. So this is very quite clear. Research lead if the email's verified as business. Okay, so this is the title of that tool and then the description of that tool. Now this one is a very big tool. you can see on the left hand side it's lengthy'cause it's doing so much. Now this rolls into how complex do our tools need to be. Now I could have quite easily have split this tool into two. So we researched the lead first, then we researched the company. But in this case, I found it easy enough to just put everything into one tool. That's what we did. so when I was mentioning before as well, just having our company information, this is the company information that I've added that is relevant for our research. So we can see here now the full name has gone in, the leads email has gone in and our agent has added, has put all of that in for us. And then we're gonna go through all of these different steps. Some of them is research, some of them is manipulating the data. some of them are LLM steps. And these particular ones now get into our LLM steps. So this is the LLM steps here. Really collecting all of that research, just doing stuff with it so that we can then generate our outputs at the end. Now our outputs at the end, there's a lot of them. Okay, so there's all of these different outputs. Now, if we actually, so again, just

Isar Meitis:

to give examples to people. We have lead, job title, lead email, company name, company revenue, company size, all these different things. These are the different outputs that this agent is going to output. After doing all the research, filtering all the filters, and packaging the information in the way that will be useful for the continuation of the process.

Ash Stearn:

Yeah, exactly right. So, like I said, this tool could have easily been split into two different tools. It didn't have to be all packed into one. but in this case, I've just put it all into one tool. just because that was just my way of thinking. Now my way of thinking is not gonna be the same as somebody else's. there's not one right or wrong way to create your tools. If it's giving you the output you want and you need, then nothing else matters. You just gotta get it working and that's it. and it doesn't matter how it looks either. so in this case, as we can see, like we've just run our tool, it's done all of the stuff that it's needed to do and this is what the outputs are. So we've got our company, LinkedIn, URL, the industry number of employees, the lead summary LinkedIn, URL, and then all this other, data company related, sorry, lead instrument. yeah, exactly. And so next our agent is gonna take all of this information, which is a lot of information, and it's then going to generate our pre-sales or report. and then what this one does, it goes through a whole different process of, again, just doing, a little bit of content research on their LinkedIn, to see if they've got any recent content that they've posted. And again, it's going to then generate our pre-sales core report, which is what we saw in that Google document sheet. Now, just to reference again, just to show it again, this is what that looks like for us.

Isar Meitis:

Yeah.

Ash Stearn:

Right. So this is now all of that research all into one thing. We've done a whole bunch of different stuff with it, and then we've generated this report. Now after it's done, what happens then, after our lead researcher has actually done all of this, it then sends all of the lead information to our CRM update agent. Now if we view the conversation here, we can actually see what information has been passed to it. Okay? So in this case, it is, this particular information that we want uploaded into our CRM. It's used only one tool, and that is just, running some different steps that are connected to my CRM that I use. And it's, capsule. Is it capsule? It's capsule I think. and then this is the information that is uploaded. Now. I've seen people obviously upload it into, more well-known CRMs like HubSpot. And within HubSpot I actually set up for a client, to have this pre-sales core report actually then uploaded into, I think it was the notes section, of that, new lead that has just been, input into their CRM. so there's various different ways that you can use this as well. This is just my particular setup that I had for me. And then that's it. And then our manager at the end of all of that will just give us a little update. the information that's been updated into your CRM, just for our reference and here's the presales core report. but what I was doing is I was actually already seeing that within the CRM, so I wasn't usually coming into the dashboard at all. I straight after a call, I would then go into my CRM, just make sure that the new lead has been uploaded into there, and then go into my Google Docs and I'll see that document that was in my Google Docs already. and then that's when I'd read it. but for my discovery call.

Isar Meitis:

That's fantastic. I want to touch on a few points to expand people's horizon beyond, yeah. this, and this by itself is incredible. So quick recap and then I'll add my 2 cents. So the recap is, relevance is one of several tools that are for non-developers to be able to create agents. And the way you do this is you create instructions for different, a, well, first of all, you map your processes, you understand what needs to happen, you define the different agents that needs to work, and you define which tasks it needs to do. And for each task, you basically need to generate a tool to allow them to do it. Then you orchestrate it just like a regular team in this particular as a manager, and then there's a researcher, and then there's the update, the CRM, kind of thing. But you can add stuff beyond that. So if you think about like the broader process in a company, the next step would be to, Do a lead score and decide what should be their next step in the CRM itself. So are they ready to buy? Do they need to be nurtured? Do, are they ready for a call with a sales agent? Like what should be the next step? And you can analyze that with an agent and update the CRM accordingly. You can even generate. The content that they need to get as the next step. So let's say you wanna send them an email, instead of just sending them a generic email from your CRM that's just automated, you now know all this stuff about them. So you can send a very personal email that's gonna be uploaded to the CRM and send from the CRM just like all your other emails. So it kept being tracked and everything else that you get from the CRM, but it's customized to this particular person based on now you want to take this to the next level. You can connect this to your note taker. So I use Fathom, but there's a gazillion other note takers. So you can grab the action information from the note taker on the actual meetings you had with them and other people from their company. And that could roll into deciding what's the next step that could go into the next piece of content that it generates and so on. And over time you add more and more of these steps.

GMT20250424-154653_Recording_gvo_1280x720:

Exactly.

Isar Meitis:

And more and more these agents. So. One of the problems, the thing people have is they look at the big picture and like, oh my God, how do I even get started? This is insane. But once you break this down, like, okay, and I'll go back to what you said because I think it's awesome. What is this thing I hate to do the most or waste the most amount of time for me or for my team? And start with that. And then, okay, what's the next annoying thing that I don't necessarily want to do? and I think it becomes interesting because that makes you decide where you're actually providing significant value. And where you're not and where you're actually providing significant value, you could still be a part of the process. You could have this send you an email and say, this is what I'm thinking of sending them as an email. Please make your suggestions and corrections and send it back to me. And then you email back the agent and the agent's gonna say, oh, awesome. And then it's gonna keep on going from there. And then you've given your input on what you talked about with this guy when you're having coffee the other day and you talked about his, him and his daughter going on the strip. And you can add that stuff that the system just doesn't know, or won't think that it's important to add. So this is where you start adding the human inputs and just taking away the tedious work that is just tedious. Like doing research is not, you're not providing any value, right? Like these tools will do better than you every single time if you created the agent, properly and it's gonna do it in 10 minutes instead of you investing three hours.

Ash Stearn:

Right.

Isar Meitis:

And so, exactly. You start, need to start thinking on how to break down the processes that are happening in your company. I shared earlier how you can do that and then start doing small steps. Build the first agent with the first three tools, tweak it, and tweak it again, and tweak it again. And you're like, okay, this is awesome. This actually does this part of the work. Then go to step two, step three, step four, and add more agents and more stuff. and this is how you take away more tedious work from more people and making them happier employees and making them actually work on stuff that matters versus doing research or updating the CRM.

Ash Stearn:

Yeah. Yeah. I think you really banked it, like you nailed it on the, right. Spot on with all of that. I think what I love about, doing this kind of stuff is that you can just gradually add to your team. you don't have to stop at just what I've shown you. Oh my goodness. The stuff I've built since that has been huge and I'm just constantly adding to it. thinking of different departments that I wanna have, and how is that all gonna be structured? And then of course, what you create, and you, right now it doesn't have a purpose or a meaning to it, but you might find a meaning and a place for it in the future. it can so easily, happen like that, And I think it's, quite creative. As much as it is technical, it is quite creative as well. in that, the core instructions, I know that I keep going on about it, but it really is when you think of like how it's supposed to work and function, it's so, language based, so be creative with that language and how you're giving it and how you're instructing it. And there are so many different ways and approaches to it, I love that it's not rigid it's not the workflow. The LLM model is the center of it and the brains of it. And that's what I love. yeah.

Isar Meitis:

Question about LLMs. There's a question from Catherine that's saying, are LLMs any good at walking you through how to build these agents and set up all the steps, or is it too new? non-tech, I built an app on the weekend, which at GPT, but basically the question is, can you go to Chachi, PT or Gemini or Claude or whatever and say, Hey, I'm working on this agent. Help me build up the core instructions as an example,

Ash Stearn:

right? So what I would do if you wanna do that, and I've done that with other tools, not so much with relevance say I, but other tools I've done, I'm like, please help me out here. Like, what am I supposed to do? you've gotta provide as much context as you can. So if the tool, for example, relevance, they have a lot of documentation. You take their documentation, you copy paste it, however that looks in a document or directly within chat GPT, for example, and give it as much context as you possibly can give it. and then ask the question, all right, I wanna build an agent for this particular process, try to be as detailed as you can in that process. Just like what you mentioned before about going through the process and the details and mapping it all out. Give that then to chat GBT and it will be able to help you, think of how you can approach building it. It might not understand the nuances to it, the little things. those are things that you will just have to learn as you go. But in terms of structure, in terms of plan, yeah, it's gonna have it all, just context. Just provide as much as you can.

Isar Meitis:

Couldn't have said it better. This was perfect. Context is everything to these tools, and the more you give them, the more, the better and more relevant the answers are going to be. the two things that I will add to what you said one thing is today, you don't necessarily have to copy and paste stuff. You can give them the URLs to the documentation. They know how to read web pages actually really well. the other thing that you can do, which I started playing with, not necessarily with amazing success, but with some success, and I think that's the future of any tool user manual, is I started using Gemini. So Gemini in their, experimental version. So if you go to. AI studio.google.com or something like that. basically their experimental environment. There's a live option and you can share your screen with it, and then you can see what's on the screen. And then you can literally just talk with your voice, say, Hey, I'm developing this agent. Here are the instructions. This is what it's doing. I need your help in writing better instructions for it to avoid the thing that it's doing. And then it's actually seize the screen to get it with you. And you're having a conversation with your own voice, just like there's a person there. Now, I think what's gonna happen over time, I think that's gonna be. The user manual. There is not gonna be a user manual. There's gonna be, okay, walk me through what I'm trying to do. And it will explain to you and highlight things on the screen live as you're doing the thing, because you're gonna have the ultimate expert there with you at any given time. Right now, Gemini is limited with what it knows about it. But again, you can tell the Gemini, okay, let me open the user manual here and read through it, and then let's try to figure this out together. what I found is when I run it for about 10, 15 minutes, it just crashes. So when I'm finally almost about to solve the problem, but I did have a few successes when it was shorter sessions and it's magical because it's just, yeah, there and it's an expert and it can help you and give you guidance and so on. It's actually looking at the screen together with you, so that's another thing that you can do. Ash, this was fantastic, like, I think as an introduction for people to the world of agents and how they work and how they're built and the intricacies of, what's required to build them. This was an awesome session. If people wanna follow you, work with you, learn from you, hire you. What are the best ways to do that?

Ash Stearn:

look, just LinkedIn. I'm very active on LinkedIn. I post quite regularly. I have, also just, opened the doors to my new community for women in AI as well, where I actually teach beginners and non-technical, people how to build AI agents with relevance ai. I only stick to relevance AI because there are so many tools out there. I'm like, no, let's just keep it less as stress free as possible. so yeah, I just really focus on relevance AI for that. But, but yeah, LinkedIn. Yeah.

Isar Meitis:

Awesome. Perfect. anybody who's joining us on LinkedIn or on Zoom, I really appreciate you. I know you have other stuff that you can do on Thursdays, early afternoon in the US or wherever they are in the world. There's usually people from all over, joining these sessions. So thank you so much for spending this hour with us. If you're just listening in your car or in your headphones or whatever, I appreciate that as well. we just crossed 250,000 downloads since this podcast was launched, which is an incredible number to me. but one of the things that we wanna do is we wanna learn what you want more and less of this podcast. So in the show notes, there's gonna be a link for you to. take a survey. It's gonna take you a minute to take the survey and it will give us feedback on what you like and don't like and will allow us to focus and do the podcast even better for each and every one of you. So I appreciate you being a listener of the show or joining us live. I really appreciate you Ash, for joining us and sharing your brilliance. and thanks everybody and have an awesome rest of your day.

Ash Stearn:

Awesome everyone. Thank you for having me. Bye.

Isar Meitis:

Absolutely. Bye-bye.

People on this episode

Head Shot

Isar Meitis

Host