Episode #22: Beyond Buzzwords: Practical AI in Manufacturing, with Radu Spineanu of Humble Ops
In this episode of US Manufacturing Today, powered by Veryable, host Matt Horine and Radu Spineanu, Founder of Humble Ops, dive into how AI is revolutionizing the manufacturing industry. Unlike abstract R&D or buzzword-heavy platforms, these AI tools are being designed for real-world application by frontline workers and managers. From the difficulties of data capture to the exciting potential of generative AI, the discussion covers the transformation of manufacturing through digitization and automation. The episode also touches on the concept of 'vibe coding' and the future prospects of AI making technological advancements feel almost magical. Radu also shares his excitement about the potential for American manufacturing to regain its edge and become more attractive than the software industry.
Links
Timestamps
- 00:00 Introduction to AI in Manufacturing
- 00:39 Challenges in Manufacturing Data
- 01:06 The Role of AI in Data Capture
- 01:21 Controversial Ideas in AI Development
- 01:56 Human Input vs. AI Capabilities
- 02:00 Excitement Around AI and Manufacturing
- 02:36 Technological Innovations and Future Prospects
- 03:20 The Magic of Rapid Iterations
Episode Transcript
Matt Horine: [00:00:00] Welcome back to US Manufacturing today, the podcast powered by Veryable where we talk with the leaders, innovators, and change makers, shaping the future of American industry, along with providing regular updates on the state of the industry, the changing landscape policies and more.
In this episode, we explore how AI is making its way into the heart of the factory, not through abstract r and d or buzzword heavy platforms, but through real tools that frontline workers and managers can actually use. Our guest brings a unique background in engineering, entrepreneurship and productivity tech.
Now through Humble Ops, he's focused on solving one of the most important and underserved problems in American manufacturing. How to extract AI driven value from the physical work of making things. Radu, welcome to the show.
Radu Spineanu: Hey man, thank you so much.
Matt Horine: Would love to dive into your journey and what you're building with Humble Ops, and you've had a fascinating journey from building developer tools to now focusing on frontline operations and industrial environments. What drew you to the world of manufacturing?
Radu Spineanu: I think what drew me to the world of manufacturing is my previous [00:01:00] company was a Universal API for E-commerce. At some point after we sold that previous company, I was fascinated by just SAP as a company and ERPs, and I was, it was a weird obsession. Like I've heard that before, but I was just curious about it because it made no sense.
The whole market made no sense. And so what I did was like, for a year and a half, I visited factories and I just asked them how do they work? I asked them about the tools they have. I talked to the people. I didn't do a shift, but I, I shadowed someone for a whole shift. It was just I learned a completely new world, which I had never seen before.
If you're in tech, it's very hard for you to imagine when someone is on the shop floor having to scan something, manufacture something, a line worker and so on. It's completely different than anything you can imagine, and I just got fascinated for it. I initially wanted to build a universal API for ERPs, which was a horrible idea.
And then I switched to like other, the other ideas, more crazier ideas, I would say.
Matt Horine: You [00:02:00] said something really interesting there about how maybe sometimes tech doesn't line up with the realities on the shop floor. Was the shop floor moment for you, what kind of sparked the idea
Radu Spineanu: For humble ops, it was a longer process. I think the moment being on the shop floor is imagine that you go to a university you've never seen before. The opposite would be. Person who is used to working on the shop floor all the time, and then they have to, they get promoted to becoming the chief advisory, whatever experience for the business and they have to go to the business thing and they are used to people cursing and being on very natural and so on.
And then you have to go in and have to and see complete this new world that functions completely differently. That was, that was me, where like I, I learned this world and I just. Connected with the people and I think I wanted to build stuff. I, I knew I was gonna spend the next 10 years and I said, I wanna spend it on this space.
Matt Horine: It really is one of the historically, probably pretty underserved by tech or slow to adoption, or there's all those [00:03:00] stigmas that are around manufacturing, but it's where most of the innovation in our economy goes on. So it's matching up reality with what expectations are. For listeners who haven't heard of Humble Ops yet, what's the core problem you're solving for manufacturers right now?
Radu Spineaunu: I think the biggest core problem is if you're manufacturers, you, lemme give you an example, right? You know, let's say what your output numbers are, but you have no idea why. Let's say a shift, the B shift from LA yesterday produce 20% more output than normal. You don't know. Which are your workers are the best or not the best.
You don't know how your impact your, the impact of your materials not arriving in time effects on order happening two months from now. And some, oftentimes someone is asking you if they can take an order rush order and they pay you a lot of money and you don't know because you have no idea about your schedule.
And in the end, this all goes back to the idea that. Right now manufacturers don't have good data and don't have clean data. So [00:04:00] there's various from people that are still on like paper checklists all the way to even people that they have their own data lake inside on-prem and doing something and they store so much data, but nobody knows how to look at it.
Nobody understands what to look at it. So we know there's this super intelligence coming that's going to be able to solve a lot of problems and it is coming. We can definitely talk about that. The core tenet of anything that needs to exist for, in order for anything to be impactful is it needs to know what's happening on the ground floor.
It needs to know the ground truth, which means very clean data. And basically the word for this is the dirty word you've heard before is like digitization. And for the most part, this was complete garbage because people would, it would be very expensive. It cost like hundreds of thousands of dollars, six months to implement.
When it was done being implemented, he would not ma match the ground level truth anymore. And then he promised all this data, but then nobody would know what to do with the data. Right? Nobody would absolutely not do [00:05:00] anything. So people treated it as like just a pet project. There's probably a laundry of a long list of, uh, digitization project, which just failed or just stopped it, right?
And people just ended up going back to Excel. And that is all changing right now for the first time in history because. You need to do, if you're a manufacturer, you need to do digitization. You need to do data capture. You need to do clean data capture, but not for you, not for the humans, not for anybody else.
You're doing it because you're gonna have this AI that is going to be able to generate all this software for you on demand, like overnight. Literally overnight. That's going to analyze all this data, right? 24 7, that it's going to be able to give you like ideas from improvements, ideas for how to do quotations better, how to move everything, basically move.
Like 10 x faster, but nothing is gonna happen if you don't start with the core idea, which is putting a foundation off of good data.
Matt Horine: You said something really important there. In the foundation of good data and a past life, I was working in a manufacturing company. We had various systems for [00:06:00] ERPs, or even on the commercial side, the CRM, you would try to make a switch.
It's back to the spreadsheets. Your data's not as fresh or which version you're working off of. ERP implementation is something that has historically been a nightmare for people to make some kind of transition because everyone knows the data is probably not clean. And then two, it's actually the mechanics of the implementation and there's a whole ecosystem out there not knocking anything.
But this is something that seems from a manufacturer's viewpoint extremely. I don't like the word disruptive, but something that is an accelerator where there's a entire ecosystem of people that implement, they're not even with the big name brands and won't put any on the spot. The name brands that are out there have this contingency of consultancy groups and others that take on long engagements.
So the startup costs are extremely high. And then to your point, you don't get the data that you're looking for, or it gives some kind of output that. It's great, but is it actionable? Is there something you can do about it or be warned in real time to take the next step or look out into the future? [00:07:00] It's mostly just capturing some type of movement.
So that kind of leads me to my next question. There's a lot of AI hype out there right now, and what makes your approach different, and how are you embedding AI into those actual workflows and not just dashboards. We have a constant joke about like, dashboards isn't giving us the insights that we want, but would love to hear your thoughts on that.
Radu Spineaunu: I would say like the most controversial thing I can say right now, it's the idea of vibe coding, but I think the idea of, if you've heard about Replicate, and it's the idea where you give a prompt to the AI and immediately starts coding things for you, and you don't have to do anything, and that is a very controversial idea.
I think what we're doing differently than anybody else is this one idea where. We are turning people from necessarily people having to think to curators. And here's what I mean by that. Let's say you have a company. You already have a system. You're building a system, right? The way, for instance, sample works is you give it a specification and in 24 hours, literally overnight, we'll deliver you, we'll deliver the first system, right?
But in 24 hours, you don't have to do anything else [00:08:00] except literally paste the document, answer some questions, and that's it. Now you come in the second day and you test what it did, and you find a bunch of problems. You tested, this didn't work you, this didn't work, this doesn't feel right. Whatever you write, I wanna change this, I wanna change this, I wanna change this.
You answer some questions, and you go and you come back. The second day you test again. You see what you did and so on. Oh, you didn't like this change. Oh, no worries. You just roll back to the previous day. You give it another problem and you try it again. This transforms you from, you have a kind of idea where you're going towards, right?
You can have an interesting direction, but then what happens is you are able to do these fast 24 hour iteration sprints, where in the end you will get exactly what you want. So imagine that you are manufacture, the current process right now is like someone has to come and that they talk to everyone. And they do this six months project and they start coding it and they write so many specifications.
Everybody reads the specification, but everybody who's read the specification knows that unless you play with it in reality, you have no idea what it's gonna look like. And this process [00:09:00] takes six months. In this new world, you can have, in the equivalent of one month, you can have 30 iterations. If you work a lot, you can have six iterations once, one month.
And you think with literally 24 hour, and imagine where like before you had to have an ERP and MES, it shattered. The original RP idea was this original idea of, or not have everything under one silo. So everything is going to be shared. They're gonna get all the knowledge, but that in reality failed and it shattered internet, em others, s qss, PL, and all the kind of thing.
And they specialized. And because of the way software had to be created before it, they had to created a one size fits all sort. And that is not necessary. Like you, you can just. Create a system that is, has that 5% of the ERP that you care about, that 5% of s that you care about, the 5% of that, but it's made specifically for you and it's made overnight.
Now, what this does is it gives you access to data, right? But once you have access to data and you have a system that can change itself, you have an AI. [00:10:00] That can go and start doing deep research. You can ask them like, why did I produce 20% more? So the AI immediately has access to all the tools of all the information, all the things can and can vibe, code, integrations into whatever.
And it says, oh, I've noted this pattern today that whenever you are ordering a part from this vendor, it's always four days late. And that basically causes all your orders to be shipped late. Here's some ideas of how you could fix this. And this is like just one example,
Matt Horine: I think you said something really important. The 5% of the system that you care about can become customizable, and even in manufacturing we talk about how much of it is customizable versus mass produced, and the solution has to fit. The product and ultimately if you've got something that's a one size fits all with a little bit of customization, and it usually revolves around how many seats do you have with some kind of enterprise software or what fields can I add, those aren't the things that most businesses that are looking for bigger and more actionable insights are actually looking for right now.
It's something much more transformative, which. Brings me to my next [00:11:00] question. In your experience, what's the biggest disconnect between traditional enterprise deployments or even some traditional AI enterprise deployments? You were talking about five coding, and what's the difference between that and what actually drives productivity inside a plant?
Radu Spineanu: Even when I talk to people right now, 90% of people say I use it to clean data. I use it to polish my sales emails. I use it to create my quoting system to answer quotes and so on. There's a couple people that are like on the other side of, part on the robotics side who are having, there's a bunch of people that have huge successes with the light south factories and, um, bridging a bunch of robotics in.
When you go into this insights, right, how can I actually run my factory better? I ask, I ask people and I ask for manufacturers like, how. Mark, do you trust the data you're gathering right now? Let's say you, you have this data, how much do you personally trust it? And the answer is like 50% in, in best case scenario.
And almost everybody I talk to either has a project to clean data or to figure [00:12:00] out how to capture more data and, and then that's just the beginning part. Because once you have this data, you have to start doing insights, then you have to have to build the act layer. How can you actually take actions based on that data?
How can we have a system that basically gives tasks to people to start going and doing and so on.
Matt Horine: Now that makes a lot of sense. And to that point, when you roll out a new ERP or you're doing some kind of process management or anything lean, the change management and implementation factor is one that it can be an obstacle mostly because people are used to doing things the way that they always have or.
Whatever attribute you wanna just ascribe to that, but how do you help operations leaders introduce AI to teams without overwhelming them or disrupting the daily flow? We hear a lot of buzz about that. Oh, AI is going to take jobs. I think that's foundationally not true. I think it enhances. The role of the operator.
You hear that AI can do something that maybe isn't that value add. We have a great phrase at Veryable from our CTO, and so I've gotta attribute the quote to him, [00:13:00] Noah Labhart. It's a AI for adults. It's something that we're not doing, something that doesn't add value, but how do you get that in the hands of the right operations leader, and how do they roll that out successfully to their teams without disrupting what's already in place?
Radu Spineanu: I think the secret is, and I think the secret with like Humble is you don't have to marry a platform like Humble. The expression is you can take a, you can take humble out for a date. It means just like Palantir doesn't agree, like humble can work on top of your existing stack. And that means you can start by automating some things that are happening on Excel.
Like I was talking to some gentleman and they were a chemical manufacturer company. And they have a certain piece of equipment, I dunno that a certain piece of material that is the secret to their success, right? And that that material is produced from, it's getting from a vendor and it's mixed by them, but the quality of them of the product is matched by the vendor.
If the vendor ships faulty material, uh, raw [00:14:00] ingredients, then the then product will never work. Right now, they have no way of tracking which vendor, because they have multiple vendors, they have no way of tracking which multiple vendors provide the, where the material from each vendor goes into which end product.
So when something breaks in the end, they're able to track it to the individual vendor that shipped up the product. So they could find, oh, usually this vendor is giving me faulty materials or something like that. It might be just because of the seasonality, or might be because of something else, and this is just, no system has been designed to do that, but that's their need.
That's the beautiful edge case of it. Using something like us, you basically just automate that processing flow and you can connect your existing RPIE as to just right, that can read data when you need it. So. It can be limited in scope in the beginning, it can be as big and as large as you can, and we are solving a real problem.
Matt Horine: That's something that makes a lot of sense and that kind of working on top of that existing stack is probably the future in a lot of ways where people aren't just plugging and unplugging. Things, and [00:15:00] you talked to a lot of those key metrics, people were looking to track things that they didn't even know that they needed.
Whereas before, it might just be on time in full, or it might just be those kind of standard KPIs. They're great, they're foundational to the business, but when you really get into it, that's where the RRI probably is and. Outside of that, and I guess doubling down on your example, so keeping it anonymous, what kind of other metric improvements have you seen, whether it is uptime, whether it's throughput, quality, just general communication.
Are there other examples that you can share?
Radu Spineanu: Sure. Let's say someone just wanted a master store scorecard and ba, basically they want to see how all their teams are performing and they're able, they wanted to see, okay, so when we're having downtime, who is on shift? What products were we making? What was the raw ingredients for this product that we were making?
And are we using this machine, whatever, all kinds of Veryables. And they wanted to see where just they wanted AI to go in and then they would collect, they build a process and procedures, so people collect this, all this [00:16:00] data. They made the integrations so humble, collect that data from automated integrations as well.
And oftentimes they want, they want the human person enter the data. They want to get it from the machine to see if it matches, and they're able to do analysis of, hey, what's happening inside of the organization? They're able to ask questions and the questions, it's the things they're interested about.
Matt Horine: No, that makes a lot of sense because it's so customizable and something that they want to see probably test that they're dipping their toe into it and seeing how they want to make those kinds of changes.
What's the human input error versus the actual data sets? Pulling back a little bit, maybe a little bit broader perspective, what excites you most about the intersection of AI and American manufacturing right now? I feel like there's a lot of buzz going on around Reindustrialization. We're seeing a lot of reshoring.
We're seeing a lot of investment. A lot of it has to do with semiconductors. A lot of it has to do with incentives and those types of things. But what if, as somebody who is at the forefront of this, where those paths cross at manufacturing and ai, what, what has you most [00:17:00] excited right now?
Radu Spineanu: I'm an immigrant, I'm a citizen and an immigrant, and I generally, I absolutely love the us.
It is my home and I just wanted to succeed and I wanted to be a healthy country. And in, in the past, let's say 10, 20 years, there is, has been an over financialization of everything and I don't think it's healthy for any country to not manufacture things. And I do believe like a lot of what happened was a lot of the core of America was affected by this.
And I'm very excited to be able to see that there's a bunch of reshoring or bringing it closer to home and people more, much more people trying to work. And I hope that this industry becomes a lot sexier than software industry. Because if you think about it, like software is this insane business model where you have like a global network effects, and it's just so easy to have these insane valuations and insane money, but, and you understand why? Because they just offer times their winner takes it all. But is it healthy to just have an industry that, a, a [00:18:00] country that basically just prints money and gets products from somewhere else?
I'm on the belief that I would, I would like to spend a bit of time, and I'm not saying there's so many amazing companies in this space, but just trying to make it easier to manufacture things, trying to. Make it, make it more enjoyable. Right. In, in, in a tiny bit that we can, right? There's so many people tackling for so many orders.
There's amazing companies. Again, Palantir, Hadian, just the name of you, right? There's so many, and I don't think, I generally don't think there's a competition between anyone because they're all trying to do the same thing is like make manufacturing a market that where. Would just be able to absorb all the investments and be able to generate returns that investors more and also make amazing products.
Because I remember when I was growing up, right? I was growing up in, when I was growing up in like early nineties, right? In Romania, we would always have this idea that Americans have the best products, they manufacture the best thing, but something would be manufactured in the us We'd be like, oh my God, did you hear it?
John has this thing that was made in America. Oh my God, he's so cool.
Matt Horine: Man, you [00:19:00] really, you really hit the nail on the head for a lot of what we see with our guests about if a country is somewhere where it's just providing services and not necessarily making things and just taking capital and not really increasing their output, the system is really on line right now where we can either continue doing what we're doing or we have to really take a step back and aggressively attack this.
It's an exciting prospect for the technological innovation side of this and the AI components of what we're seeing. You sift down through all the buzzwords and you can see how it can be a, a productivity enhancer and how it makes people genuinely better operators and faster at their roles, and ultimately where we can be at a place where we make things again.
We've talked to a lot of guests who say, we do want this kind of manufacturing. We don't. I think the foolproof answer is we want all of the manufacturing back that we can get. I always joke about the toaster. Somebody said they don't want to make toasters in America, and it's like, you can find a great way to make a toaster.
Let's find a way to make a toaster in America. And I think that's generally the sentiment that [00:20:00] we get. But along those same lines, it's a pivotal moment where people are nervous about this because they've recognized the problem. We don't make a lot of our own medicines. We don't make a lot of our own steel.
Our shipbuilding capacity is abysmal compared to the rest of the world. You take that and extrapolate it over any other type of industry, the numbers aren't looking as great either. So people get a little nervous about the AI component, but I think there's probably a really good opportunity here for you to define what's a myth or misconception you keep running into when talking to manufacturers about this and how do you sort it out or dispel the myth?
Radu Spineanu: I think the biggest myth or misconception that I, when I've talked to people about AI is they still don't understand how good it's gotten. Which means a lot of the conversations that I've had are either I'm using it to like clean data, write emails, et cetera, or they're saying, oh, I used the chat GPT or whatever, six months ago, and it hallucinated some things.
And it's very hard to explain that how fast [00:21:00] this world is where. Imagine GPT three I think was released in 2020. Then you have Chat GPT in 2022, and then we are here right now where I think with the latest releases, right, Gemini and Claude, and now we're gonna have GPT five coming probably very soon. I use them every day and I, we use it for very extreme things and I'm also in groups and people that use them for very extreme.
Things and we internally are seeing using every day we, we are having existential crisis all the time about how good it is because we cannot see what it cannot do. And people go in, they type on Chat GPT, and they ask you to do some things, but they don't understand that the fundamental difference is not that they can do that is the fact that you can have a thousand ChaT GPTs working in parallel communicating.
Making decisions, understanding various things like for the first time in history, we are automating knowledge, intelligence, and a lot of these things are actually smarter, already smarter than most human beings. [00:22:00] And, but imagine not having one. It's not about when you talk to one, imagine that you have in the future we're gonna have 10,000, and you're gonna have 10,000.
You're gonna talk amongst themselves, do research, argue, fight, validate, deterministically, non-deterministic. Who knows? They're gonna offer something, right? That is basically the equivalent of you having 10,000 people that were trained to do something to solve a problem for you. Whether that's software generation or that's research, whether that's designing a new plan, whether that's something, and the thing that people don't understand is how good it's gotten and how fast we're moving and.
This idea that it's not about a chat bot, it's about it in the backend. You can start connecting all these intelligences together and they can have. Real conversations and talk and brainstorm, and it's just getting so much better, so much faster.
Matt Horine: A lot of people are still using the frame of references, asking it almost like a search engine or generating AI art or doing those kinds of things.
It's that real kind of consumer mindset where you're a [00:23:00] user of it and. Maybe how it helps you do some things. And I've seen people use it for budgeting and I've seen people use it for things that are tasked but not really talking to each other or thinking about that infinite capacity that you can generate off of building those types of models and prompts.
So that's, that kind of leads into where do you see the greatest opportunities over the next two years, let's call it, because it is moving so fast, the greatest opportunities over the next two years for Humble Ops.
The biggest opportunities
Radu Spineanu: for us is I think we have to be patient. When we, when I started this company, I knew it was gonna take 10 years to build it.
I think outside of this big PR of which we don't have the money to, so the end is coming at the guy with the bell is we have to digitize now. Or at some point, I think there's gonna be a lot of technology coming that's gonna feel more and more like magic and we wanna be part of that technology that feels like magic.
And we, as I mentioned, we have to start at the beginning, right? Everybody I talk like this is, I'm gonna say something that everybody knows. Garbage in, garbage out. If you give AI garbage in, it's gonna give you garbage out. Everybody knows that, right? So. [00:24:00] My first job is to convince people, is like, Hey, right now you have to digitize.
You have to put your processes in place. You have to put your procedures in place. You have to make sure like people use this. And again, you are not doing this for another human being. The people on the floor are not doing this for another human being. Who's not gonna look at the data, who's not gonna care, who's not gonna change something?
You are doing it because you're putting a foundation, right? You're putting a foundation on which you will be able to collect Digital twin, the buzzword. Use now, but you're gonna collect the actual data, including what the human people, like the people do on the shop floor. And then you'll allow it to analyze this data, to distill this data, to figure out what's happening with this data.
But then you also wanna allow it to push back, meaning to push tasks back to the people. Hey, try to doing this, try to doing that. Oh, this is wrong. Here's, you should try to fix it like that. And it's honest. Oh, I talk to people and they, when they talk to high level people, it's always, I don't have visibility.
I don't know what to do. I wanna know what's happening inside the company, but then it's [00:25:00] okay. Like what? Where's the raw data for this? When you can get analysis? Oh, it's on paper. Or it's like someone enters it manually on Excel and so we, you have to start at the beginning. And the beginning is the digitization.
It's simple as that. It's like you have to get rid of paper checklists
Matt Horine: And in manufacturing it's either, sometimes it's on paper and some people. People have it up top locked in their head through tribal knowledge or historical knowledge, and they're close to retirement. It's something we hear about all the time.
So I'd say that's the one solid piece of advice you would give anyone in manufacturing right now looking into ai, it's you have to do this because of the pace and acceleration that's happening and it's a good first step to digitize.
Radu Spineanu: Yeah, I mean that's the digitization is, it doesn't matter. Humble is the company that makes it, or there's gonna be like Palantir, there's gonna be like Microsoft has whatever, right?
It doesn't matter, like the main thing you have to start doing. Is you have to, you have to start collecting the information and telling the AI what you're actually doing. You not only collect the data, but you have to tell AI what they're doing. And we're seeing this and a certain degree in [00:26:00] engineering as well, where.
Part of our job is we write algorithm. We tell the AI to do something for the algorithm, but to a certain degree, what actually makes a lot of difference, and when you describe what the algorithm does to the ai, immediately gives better recommendation. So just giving the data oftentimes is not enough.
That's why you also need to document the process and the procedures. You if you don't know them. You shouldn't wait until something is perfect because it's never gonna be perfect, because now, right now you can do these 24 hour iterations, which are not possible. That's the secret insight where even if you don't know how to do it, you write it as best you can.
You put it in front of the people. Use this. Does this match your reality? The guy says, the person says, no, wait. How is it d? And you just record a call. Like you don't even have to take notes. You record a call, you literally give the transcript to Humble and it says, okay, change it based on this. It goes and does it.
Matt Horine: Yeah. The world of possibilities is certainly expanding rapidly. You mentioned if they don't go to humble ops, but I would like them to check you out. So if our listeners wanted to find out more about Humble Ops, where would they go and where would you recommend they get started?
Radu Spineanu: [00:27:00] Right now our website is quite out of date, so we're a stealth company and the simplest way is they can either email me and I'm sure they're gonna have my email somewhere. I'm me on LinkedIn. I'm gonna comment if you post this on LinkedIn, I'm gonna be one of those comments, so like it's another way.
Matt Horine: Absolutely. No, we'll make sure to tag you. Radu. Thank you very much for joining us today.
Radu Spineanu: Thank you for having me. It means a lot to me.
Matt Horine: To stay ahead of the curve and to help plan your strategy, please check out our [00:26:00] website at www.veryableops.com and under the resources section titled Trump 2.0, where you can see the framework around upcoming policies and how it will impact you and your business. If you're on socials, give us a follow on LinkedIn, X, formerly Twitter, and Instagram. And if you're enjoying the podcast, please feel free to follow the show on Apple Podcasts, Spotify, or YouTube, and leave us a rating and don't forget to subscribe. Thank you again for joining us and learning more about how you can make your way.