240 | AI Builder’s Playbook: Turn Your Idea Into a Working Solution in few hours with Pietro Montaldo

0:00Hello, and welcome to another live episode of the leveraging AI podcast. the podcast that shares practical, ethical ways to leverage AI to improve efficiency, grow your business, and advance your career. This is Isar Maitis, your host, and I am thrilled about today's episode, and you will be as well. And I'm thrilled for several different reasons. One, we're starting to go back to doing lives. We used to do this every single week, and now I've been doing so many company-specific workshops that have been traveling literally every single week in the last, I don't know, six to eight weeks. It seems like forever.

0:34So we were not able to do this. And I still have a little bit of travel coming in the next few weeks. But after that, hopefully we'll get back to doing this every single week. So I'm excited because we're doing this live. But I'm a lot more excited because of today's topic. Because today's topic is going to solve some of the biggest problems that you are facing and that everybody's facing right now, which is deciding how to implement with AI the solutions that you need for your business. So we all have several different steps in our AI journey. Step number one is we don't have a clue. And then you take my course or somebody else's course and you start figuring it out.

1:10Step number two is like, oh my God, what do I do now? And then you start coming up with ideas on how to implement AI in your business. But there could be 17 different ways to implement the thing that you have as an idea. Do you want to create a chatbot? Do you want to create an add-on to your website? Do you want to create an automation? Do you want to build it in NA10 or in Make? Do you want to create a custom GPT? Does it need to have a video? Like all these kinds of things. and it's a mess for most people because most people approach this for the first time versus our guest today that does this for a living and hence he has a incredibly well-defined decision tree a process on going from idea to evaluating different options to testing the different options to implementing the solution and actually using it because that's what he does for a living so So Pietro, our guest today, is one of the co-founders of Enforce AI, which is a company that does exactly this.

2:06They help companies implement AI solutions across the board, different kinds, different levels of complications, and so on. And before that, in his background, he has a really interesting background because he's done, well, more or less everything, which is actually really, really awesome. So he's a serial entrepreneur. where he had several business on his own that he started, but he also had roles as a product manager, which means he really understands product and defining solution, but he's also been an analyst and an investor. And so he also understands how to be very practical in looking at data and figuring out what it means, which really makes him the perfect person to now build AI automations for companies and to explain us exactly how to follow this process.

2:47Now we're going to start on a very high level and go through the steps, and we're also going to show you examples of the different implementations. So this is going to be extremely valuable and hence, I'm really excited to welcome Pietro to the show. Pietro, welcome to Leveraging AI.

3:04In the next few years, AI technology will change our world dramatically. Whether you are a business executive trying to catapult your business forward, or just somebody who refuses to be left behind and want to advance your career, this is the show for you. I'm your host, Isar Maitis, a serial entrepreneur and an AI enthusiast. You'll hear invaluable practical tips from innovative business leaders, AI practitioners, and some of the brightest AI minds in our world today on how you can leverage AI in ethical ways to advance your career and grow your business.

3:44Thank you so much, Isa. That was such a nice introduction. Thank you for having me. Yes, I didn't say that joking. I'm really, really excited about this session. I've been following you for a while and I know how amazing all your content and how practical your content is. And again, I'll say one more thing that I really like about you. And I think that's going to help a lot of people understand. I think you're the ultimate tinkerer, right? I think you probably do some of this for fun, right? Just try stuff. That's how I started, right? I started because I wanted to learn and then i was like okay while i learn i might as well make something with it right which is definitely my favorite my favorite job i've ever had you know i literally pay myself to experiment things which which is beautiful and one thing which i want to add to your uh to your introduction is one thing which you might have noticed that that brings together all of the things which i've done as none of them was technical right so i'm purely 100 non-technical So the approach you see today as an approach which is especially beneficial for somebody non-technical, also for someone technical which has never played with the eye, it might be very interesting.

4:49But I think it's mostly interesting for non-technical people. And on the side of our business where we develop complex solutions, what I do in my day-to-day is literally going through this process of experimenting with ideas, but with the goal to share it with a non-technical audience. right so what we are talking today is basically how i go from okay i had a very cool idea when i was in the shower while i was working for a client how do i turn into something that can inspire people right i see my job a lot as giving people inspiration of what they can build and giving them a first concrete step on how they can turn it into reality and i do it with linkedin tutorials and with youtube these are the two channels where i work the most yeah fantastic and again i think so many people are getting there, right?

5:38I think we're going now from the stage of people don't have a clue about AI to a lot of people knowing about AI, but don't really know how to materialize their ideas into practicality. And that's why I'm really excited about this. So I'll give you the stage and just lead us through the process. Perfect. Okay. I'm going to go ahead and shut my screen for the people who are live, but I'm going to drive you through every single steps in any case. Yeah. And those of us who are joining us live, first of all, thank you so much. We're excited to have you. I know we haven't been consistent in doing this, but I appreciate the fact that you're here.

6:10If you have any questions, please write them in the chat, either on the Zoom chat or on the LinkedIn chat. And if you are not joining us live, then I suggest you try to join us live next time. You can be on our emailing list and we're going to send you invites to every time we do these lives so you can be aware of exactly when they're happening and be able to join us. We do this almost every Thursday. Again, if I'm not traveling, it's every Thursday at noon Eastern time so you can join us this way. But again, back to you. Absolutely. So where do I start? Everything starts from idea. One of the coolest thing and one of the reasons why I'm most excited about AI is that I am not limited anymore to what I can build.

6:49Because no matter what I'm thinking about, what my idea is, I know there is a tool that allows me to do it without bringing on our engineering team, right? So one of the goals which I have whenever I build something for myself and for sharing with the audience and for inspiring people is I want to be completely independent from the technical team. I don't want to bring anybody from our technical team on board. I am focused on building micro-apps, micro-solution, which helps you with your day-to-day, mostly business life. What does it mean, right? What are some of the solutions, just to give you an idea of what we are talking about?

7:21One could be a screener for Reddit posts, which are interesting for your company, and a draft of the SEO-optimized comment, which you can leave on that specific post. Another one could be an agent that on a weekly basis go and scrape all of the ads which your competitors have done on Facebook and Instagram and send your report of the ads that have worked the best for your competitor. Another option could be an agent to create content which respects your content style. So it's a simple micro solution which can really solve one task of your day-to-day or can enable you to add an additional task which before you weren't able to.

7:55For example, a lot of people before weren't able to post on LinkedIn. it, right? And why do I think that LinkedIn has grown so much in the last six to eight months? It's because a lot of people are empowered to do it without so much effort, right? Because they can start from an idea, from a call they had, and turn it into a proper piece of content they can share, which literally has actual insight, right? You don't need to make it up with AI, and that's not going to add value or bring you anything. But you can turn your real insight and what you have into something that works very quickly, right?

8:23So everything starts from idea. when I have an idea of something I want to build in my specific case I look it through the lenses of okay I want to turn it into something I want to educate and inspire people with so I look for for some specific characteristics for me an idea has to be appealing for a very large amount of people right because that's that's what goes on social media right this won't be the same if for example you are studying an idea to apply it in your day-to-day you just need to look for a pain point and you can build an idea for me an idea also has to be appealing for a large amount of business people.

8:54The second thing, and probably this is the most important one, it should have been first actually, is it has to be truly useful, right? I'm not going to publish something which sells smoke. I'm going to publish something where a person tries and says, oh, this guy has done a very cool, simple thing. Let me see what else he can do, right? And then they book a call with me and we turn those people into a customer so much easier than anybody else. The third thing, it has to generally be relevant for sales, marketing, operational, or personal productivity. These are the four largest buckets of non-technical people.

9:22And also, luckily, where companies are happier to invest because it's the one that have the highest impact on the top line. And then the other thing which I'm looking for is solutions must require only simple integrations. I can explain in under a minute. I am not looking for complex integration. For example, if you're working with Salesforce and you have a very personalized sense for your company, I would recommend, okay, go and get an agency that helps you set up a proper automation or work directly with Salesforce. I'm looking for things which can be integrated quickly. I can explain it for a non-technical person and he can understand and replicate it.

9:56And I usually aim for a maximum of two integration per solution. More than that, it becomes a more complex solution, which I can describe for you, but I would want to build it for you. I would want to bring an engineer to make sure it can go into production. and the fourth rule and this is something very personal for me every solution should take less than four hours to beat right so if it has been more than four hours i put it in a bucket i hope one day i'll have someone to help me in this process and i'll get them out and i'll invest in those but i want to push out minimum two new things per week that's how you have you can stay up to date and this and with how fast the world move so i can only apply for our per solution And this is - Fantastic.

10:34Two quick questions, Pietro. One is, how do you document your ideas? Do you voice type them? Do you write them in a document? Do you put a voice note on your phone? So that's question number one. And then I'll ask you question number two. Yeah. So when I'm on my laptop, and most of my idea unfortunately comes when I'm on my laptop, because I'm on my laptop approximately, I don't know, 10 hours or 15 hours per day minimum, I have an idea database in Notion, and I just add them there. I classify if it's a lead magnet, a build, or idea for another post. But I have a nicely built knowledge database where I put all of these.

11:09So you do it in Notion, so anybody can do this in Monday, Jira, Asana, ClickUp, like any of those? Anything. Literally, I had it in ClickUp. I moved it to Notion. Absolutely the same thing. If there's a Kanban, perfect. You can do it there. You don't need anything else. Second question is, you talked about the four-hour limit. How do you guesstimate how long it's going to take you? in general the rules which i described are thought for it to be within those four hours right if he has less into connection if it's of limited complication it generally should respect those and also i get better i like to think i get better over time so now i rarely invest time when i get to the four hour mark and i'm like i couldn't finish right i don't i don't remember the last time it happened and at times it's just because maybe there are still some things that require engineers to come on board and maybe i should just call an engineer or gave up before

11:59okay okay so moving on the idea phase is the coolest one and you can have them wherever you want and when you have them just note them down if you are around i use a note on my iphone and then in the morning i have a blocker to transfer everything i thought you read the night to my laptop right now the question is and this is the coolest part of this workshop i think and i've never done this so bear with me i like eyes are happy make this clear for everyone how do i turn this into a working solution. So all of us in general use AI in three main ways. We use them in copilot, which is when AI is embedded within another product that helps you make better use of that specific product.

12:36I'll show you a couple of copilot later. We use via chatbot, for example, like use chatGPT or cloud or any other LLM, and we use them as agent or workflow. In general, when we talk about agent or workflow, we are talking about something that can do work on your behalf right yeah i like to think of two main characteristics of agent to distinguish what what is an agent and what is just a chat right an agent is proactive so it can start working on his own for example if he every monday starts working in a specific tab on his own without me going down and telling okay this would be classified as agent and also if it's not a single task but it's repeated and he performs a set of actions repeatedly and proactively, this also means for me it's an agent.

13:23Because basically the two things work together. He works on his own and he repeats the action on his own based on when he receives a certain input. It can be time, it can be an email receive, it can be, I don't know, a task completed on OSHA, right? Yeah. Any trigger of something that happens in your tech stack. Exactly. Any trigger that's something that happens. In general, and I'm not sure if theory of agents also confirms this, but when the tool is integrated with my tech stack, I also consider that to be an agent, right? If it's integrated and if it's proactive, it's integrated and it's repeated, it becomes an agent.

13:56We're not talking anymore about a chatbot, but we're talking about something that is independent and can perform work on your behalf, right? And this is the biggest and best revolution of AI, right? You literally can free up time because an agent can go and do work on your behalf. When you sit at the desk and I already have the idea ready for my LinkedIn post, or I already have my YouTube script draft. So once I respond to this question, I know which direction I want to go. If I've responded, okay, yes, this is not an agent because it doesn't respect its specificity, it's probably false under the chatbot idea.

14:31When I have a chatbot, when I know it's a chatbot, I have a series of choices I can do, right? and the most important one is what interface do I want to do to this chatbot a chatbot can be something very useful it might sound diminishing now that I mention it this way but a chatbot can be incredibly useful for example I can give all of my knowledge about how to write a good hook for a LinkedIn post and put it into a chatbot and the chatbot only performs that function that function is incredibly helpful I've spent I don't know how much time crafting the way I do the hooks and now I have a chatbot will help me get to a hook so much faster than before.

15:06Yeah. Yeah, to put things in perspective, and we're talking about chatbots, we're not just talking about free conversation or open conversation, but we're talking about tools like a custom GPT or a cloud project or an assistant by OpenAI that can then work. So it's still a chat interface that is a dedicated chat interface to one of the language models, but it can have instructions and can have a knowledge base and it can know how to follow things. I probably have 25 or more either custom GPTs or cloud projects that we use every single week in the company to do stuff. So yes, there's no reason to...

15:47I'll say something else. Anything you can solve on simpler solutions, go to the simpler solutions. So don't build a more complex solution if the simpler solution can do that. And there's a lot of things that you can build with these chatbots today. Completely agree, right? which is very important to go through this step, right? Is this an agent or not? If it's not an agent, there's no need to make it one, right? It happened to me a lot of times that I complicated things very much. I use different models, I use different steps just to realize that I have a knowledge base and I wanted to give it to an interface which can think based on that knowledge base and give me a response, right?

16:20That makes a chat, right? And you can also make a chat with fans, right? Like you mentioned, it can be a simple chat and it can be like, I'm opening up here on the screen for the people who are here live and we'll see this on YouTube. On ChatGPT, it seems absurd, but most people don't know that you can create a chatbot which you can share externally within ChatGPT, right? This is called GPTs, right? So if you go to this section, these are all GPTs, which I either rent from someone else. You don't have to pay, but you use them. And somebody else has built someone else who has knowledge on the space.

16:49For example, this YouTube script generator, it's not made by me. It's very good. And I use it a lot. Actually, this was made by me, but this was another one, which wasn't made by me. You can see here that this was created by Vaxygen SM, for example, and creates really good hooks. But then I have some other GPT's bots, which I have created myself. For example, this is a Reddit blog post architect, which works really well. I've put everything I know inside this. And if I go here, for example, and I go to add this GPT, I can see what is the configuration I gave to this specific GPT. And I can also edit it.

17:26And I edit it with a copilot here. For example, here, that's a text base that helps me get to a better version of the GPT than what I would do just typing myself. So this is what we were saying before, a copilot. AI embedded in another AI product that helps me get to better results first. So this, I have done many, many posts, and some were very successful, where the only thing I did is I put my knowledge into a GPT. I tested it for a week. If it was nice and I thought it was giving me nice results, I published and I tell people, okay, this is, chatbot does. x, y, z, I tested it for x week, these are the results, see one result here, and I always paste the result.

18:03And this is something that people are very interested with, because they might not have the time, the bandwidth, the knowledge, or three days to invest in researching how to make a really good Reddit post, right? So if you have a specific knowledge about something, you codify it, you put it in a chatbot, and you either sell it or you give it away for whatever goal you want as a lead magnet or to generate leads, it becomes something very powerful, and also a simpler possible solution you can create for yourself. So to someone who has never created anything with AI, I would tell start by creating GPTs or cloud projects.

18:35This will be on any LLM. You go, you create a personalized project within the LLM and you have a product which works, create value. You can sell or you can use internally, pass it to your team, your teammate. So you work best and it takes you a morning if you already have the knowledge. Yeah, sometimes less. So a great, great concept here. I want to add two cents. One for the people like, oh, how do you get somebody else's GPTs? If you just open the regular ChatGPT interface, the first thing in the GPT section is called Explore. And if you click on that, then it will take you to what's called the GPT store, even though it's not really a store because you don't pay for anything.

19:11It's a free store. And then you can just either go to the different sections. They have productivity and research and analysis and data and marketing and like all these kind of things. Or there's a search bar on top and you can literally just explain in English what you're looking for. and it's going to show you they now have multiple millions of these gpts that other people created some are great some are not so great but you have just to test them out and see how they work so that's on that one aspect the other aspect is on the way i teach people on how to build those custom gpts or projects i always tell people start with a regular chat so don't try to build a gpt start with a regular chat explain to chat gpt or claude if you're in claude what you want to do like this is what i'm trying to do this is going to be the input this needs to be the output here's the data that I have.

19:50Here's the knowledge that I have. And work within the chat until you get to the outcome that you want. So basically go back and forth and back and forth. Try to write, if I'm going to follow Pietro's example, I want to write great Reddit posts. So write it again and again and again until it works correctly. And once it works correctly, basically tell ChatGPT or Claude, okay, I want to turn this into a custom GPT. Please write the instructions for me. These are the files that are going to be in the knowledge base. Please write the instructions. And it writes incredible instructions, way better than I can write on my own.

20:20And it's going to be based on your back and forth and trial and error that you've done in that specific chat. So this is probably the fastest and the most effective way to create good custom GPTs, projects, whatever you want to call them. I completely agree. And then if you want to bring it one step further and you want to give a chatbot or a very simple app a face, which is nicer than just a chat, you can do it using lovable or base 44 or bolt so what are these these are web builder and app builder which allow you to create an app just with simple text right so i have some knowledge again i have it codified and i want to create an app that tells let's make a concrete example there's something i built a few weeks ago and it's a tax saving calculator when you move to another country in europe right i collected out of knowledge because i had done a lot of research because I wanted to move to a better tax location.

21:09I codified the knowledge. And now I have a chatbot and an interface, which is very simple, UX and UI, that helps me understand and engage with people better. So, okay, I do a chatbot first. I'll do a GPT project. I'll do a chat that helps me with this. I got a very good result. Let me turn it into something a bit more visually appealing. And then you bring it on Lovable on Base44, which is what you're seeing right now today. And I turn it into something a bit nicer. This is another step. And it's more of a marketing step because the very important thing is the knowledge. But it's, I have built it.

21:42Okay, I liked it a lot. Let me try and bring it one step further. At least I see it that way. You can also start directly with Base44 if your goal is to rebuild your website, for example. Yeah, I agree 100%, right? So again, for those of you who don't know Lovable, Base44, Replit, there's a bunch of others. There are basically what's called vibe coding tools, which means they can write code and do literally anything you want, but they're mostly used to create applications. And you can start with somewhere else, like a regular chat or a chatbot or a something, and then give it all of that and it will build the application for you.

22:19And then it can help you even in the design and look and feel. If you want to know more about the overall flow, I created an episode not too long ago. It was episode 216 that was called the five levels of AI, what every business leader needs to know. And it basically starts with a chat to then these more advanced like custom GPTs from there to building automation, from there to vibe coding. So if you want to dive a little deeper into that, you can go and check out episode 116. But to summarize what we've said so far is we said, okay, if something is not a chat bot, meaning, sorry, if something is not an agent, meaning it doesn't need to take action and it doesn't need to be proactive, it just needs to follow a flow based on the knowledge that I have.

22:57We went down that path that can either be used inside of ChatGPT or build a wrapper around it with like a vibe coding tool. So now I assume we're going to go to, okay, if we do need an agent, what do we do then? Correct, correct. And I love how you remember the exact episode in which you talked about that specific topic. That's some level of commitment. Thank you. Now let's go to, we respond yes to the question where we, before we have said, is it proactive? Is it repeated? Is it integrated? If you respond yes to this question, you probably want to build an agent, which is a slightly complex system, more complex system.

23:30If we have responded to this, what I think about next is, is it something that someone has already productized? And is that product really good? At times, myself, I really like to be things from scratch, right? So I have to fight this urge to be different scratch. And I think it's very useful in general to look at what's already out there, right? There might be people who have already productized this to a point where it works incredibly well, right? Some example of this. I don't know. sending call emails, it's a huge market. And some people have been starting to build that in the last six years, instantly, smartly, for example.

24:08They have an incredible amount of knowledge and they build really powerful agents for doing that, right? So rebuilding those, you would most likely build something which has super power. And also, these solutions are not incredibly expensive. Another question I ask myself, if somebody has built it and it's really nice, is it really expensive? if it's really expensive can I build something which can do the same thing or very very similar for much much cheaper so that people can basically do it themselves if the question is yes I will still go ahead and build it if the question is the product doesn't exist I will go ahead and build it if the question is yes there is a great product and it's cheap and people know it or don't know it yet even better but they can access it very easily I will go ahead try that product and probably showcase six tutorials about that product right when i have an idea and i think it's cool and i think it's appealing to people i want to just give up it's not like because i can't build it myself i won't create a resource in it i still think it's cool to recommend people to go somewhere where there's already something existing to recommend another team was done a really really good job about it 100 quick question about this where do you search like do you just go on google Do you go on Reddit?