By Jennifer Dubose

GroundControl: Turning First Articles From a Bottleneck Into an Advantage | Ep. 536

Six jobs waiting on a first article means six spindles sitting still. That is the math Mike lives with at Hill, running around 58 jobs a day where nearly every one needs an internal or external first article before parts can ship. When inspection becomes the constraint, good parts pile up behind paperwork, and paperwork does not make chips.

This week we sat down with Mehul Shah, co-founder of Ground Control, a company building what he calls mission critical software for highly regulated manufacturing. His thesis is simple. Aerospace, defense, medical, and semiconductor shops are drowning in Excel and pen-and-paper workflows, not because the people are behind the times, but because software was never really built for the shop floor or for the weight of ITAR, DFARS, and CMMC.

Ground Control started with one hairy, everyday problem: first article inspection reports. We get into how their software reads decades-old, hand-drawn prints that trip up ordinary OCR, balloons the drawing, fills out the AS9102, and pushes clean data straight into ProShop with a single click. It is the kind of automation that turns a shop’s worst bottleneck into a task almost anyone on the team can run.

We also talk about why now is the moment for this, from reshoring and record defense budgets to the labor math that says we cannot hire our way through the coming surge. Mehul shares how a 15-person team just earned FedRAMP certification, why AI lets them do the work of a company twice their size, and what their new contract review product does to catch a buyer’s price change buried in a six-figure PO before it becomes a costly mistake.

If your inspectors are your constraint and your smartest people are stuck on manual data entry, this one is for you.

What’s Covered in this Episode

  • (1:02) Meet Mehul Shah, co-founder of Ground Control: Why Mehul chose manufacturing
  • (5:20) Why manufacturing runs on Excel, and the real reason software never fit the shop floor
  • (6:04) ITAR, DFARS, CMMC, CUI, and the infrastructure tax on regulated software
  • (9:07) Mission critical software and the three tests Ground Control uses to pick a workflow
  • (10:43) Invest in yourself first with ProShop ERP
  • (12:19) Finding the wedge, and why first article inspection was the obvious first product
  • (13:49) How first article really works at Hill, and why it becomes the daily bottleneck
  • (15:53) Why the FAI is critical, process verification before you can ship and collect
  • (16:35) Why now, defense budgets, the SpaceX IPO, the space race, and reshoring
  • (18:29) LLMs let manufacturers build in-house, and when it still makes sense to buy
  • (19:30) Build on standards and a solid data heartbeat, then augment with AI
  • (21:07) A rocket ship, 200-plus customers in two years, including five public companies
  • (22:18) Growing into demand and the hard part of turning down the wrong work
  • (23:39) The seven habits of highly effective work holding with SMW Autoblok
  • (24:22) Avoiding the solution looking for a problem, and how the idea took shape
  • (25:53) Building GroundControl from the ground up and becoming compliant
  • (28:07) The importance of trade shows and building in-person relationships
  • (31:55) The story behind the name Ground Control (and a nod to Major Tom)
  • (35:20) Turn AI buzzwords into outcomes at the IMTS Industrial AI Conference
  • (36:06) Developing out the GroundControl software and finding their MVP
  • (39:32) The ProShop integration: One click from ballooning to master process control
  • (42:39) AI will not replace you, but the person using AI might
  • (46:30) Introducing contract review and catching PO errors
  • (49:37) The visitor check-in system (ITAR sign-in without the 20-year-old paper pad)
  • (52:16) What Mehul is most excited about, FedRAMP, and the road to IMTS

Resources Mentioned

Read the Full Episode Transcript

[0:00] What’s up, Mike? What’s going on, buddy? I like your shirt. Thank you. Got a Rep Hill. Hey, I just want you to know I’ve got air control covered. If you can cover a sea control, we’ll bring in ground control in a few minutes. We’ll have the whole thing controlled. Clever. You’re so clever. All right. So then they’ll play the music. If the sound of a machine tool, removing metal gets your blood pumping, then you are metal working nation. This is Making Chips, where we talk all things metalworking, engineering and design, production and tooling combined with business best practices, technology, marketing, news and new media for manufacturing professionals. Here are your hosts. Let’s make some chips. Mike, we got a good guest here today. We do. Yeah, we, uh, we’ve got, as you mentioned, we’ve got ground control covered here. So in the air, I want to hear the story behind it. I know I always like nerd out on how they got to the name, like by the, you know, whenever we bring in someone who’s got a business and it’s got a clever name, it’s like, well, what, what caused you to come up with that name for it? So Mahul, is it Shah, Mahul Shah? Yeah. Yeah. Uh, welcome.

[1:27] And you are the, are you the founder? Yes. I’m one of the co-founders. The innovator, the creator. Cool. I, I’ve held all the hats. I’ve written the software for the programs. I’ve made sales. I’ve done marketing. I’ve been the janitor. I’ve done it all. Um, but it’s great. Thanks for having me, Mike, Nick. Good to see you both again. Yeah. Good to see you. Hey, just so you know, I’ve done a lot of the Claude code behind Making Chips too, you know, I’ve, I’ve, I’ve coded our whole origin, so we share that. I’m a, I’m a coder. That’s great. And you know, I actually just saw a stat the other day that said, it’s like, we’re still very early. Only 2% of households in the U.S. have used AI. Wow. Only 2%. As much as you hear it everywhere, only 2% have actually used Chachibu. Does that seem low, Mike? I mean, my kids use Chachibu. It seems incredibly low, but, but maybe, maybe deliberately, right? I mean, I, I think they, there’s probably 98% have used it and didn’t know it. Yeah. Um, yeah, you can’t help it. Yeah. That does seem remarkably low that, you know, 2% have gone out purposely and like asked Chachibu a question or something. You know, I’ve been really focused on marketing in our business this week and just doing all sorts of research and stuff and the amount of people like who have zero click searches now, you know, so you’re searching and you say, Hey, how much does it cost to get a chip conveyor? Or how much does it cost to automate my machine tool or something like that? And then you get the AI answer.

[2:52] And then for a lot of folks, that’s sufficient. They don’t need to go to a website. Right. So, I mean, there’s an example of how you’re kind of using AI without even realizing you’re using AI. Yeah. I mean, I’ve, I mean, yeah, Lord knows how many zero click searches I do now. I mean, cause that pops up. It’s like, yeah, okay. That’s about what I thought and move on. Yeah. So, Hey, let’s get into this, the backstory, the origin story of Mahul and ground control. And at some point in the story, you got to tell us the, the name and how you came up with the name. Uh, no, absolutely. There’s, there’s a, there’s a good story behind it. So what brought you into this world in the first place? It getting you into the metalworking nation and this world we live in. Yeah. So I actually, I mean, grew up a lot around manufacturing. My, my family has been in manufacturing, uh, electronics manufacturing specifically for the last 25, 30 years or so. My dad had a business and then, you know, grew on to have several businesses. I mean, while I took a different turn during college and decided to study computer science, electrical engineering, and then I got my master’s in machine learning. And this is back in 2016. So a decade ago. So there was no chat GPT. There was no, it was all called back then. It was called deep learning. But I got a lot of really good experience in how to actually implement AI. And I went to go work in a, you know, big tech is what they call it in Silicon Valley.

[4:15] So I worked at Amazon. I worked at Yahoo. And I really just felt like a cog in the wheel. You know, you’re working, you have a team of 30 people, nothing against it, but I was on a team of 30 people working on, you know, the color of a button. And I was like, Oh, meanwhile, I’m looking at, you know, coming home, I visit my parents and I’m like, Oh, they’re doing all this cool stuff. They’re building PCBs for, you know, Tesla, Apple, Google lock, you name it. And so I eventually decided to take the leap and join the family business. And this is kind of unique. I’m sure you both can relate to this. There’s not a lot of software engineers in manufacturing. Unfortunately, there in general is a shortage of supply of software engineers. And so all the money people get attracted to, you know, working at Google and Apple and Facebook. So it was a kind of a unique opportunity to, Hey, let’s go build some real software for our own businesses and see where we can take this thing. And so, you know, the businesses continue to grow. I built a lot of internal software, everything from a customer portal to an operations dashboard to, you know, data entry automation. And through that experience, I learned, you know, a lot of interesting things. And one of the biggest things I learned is why there isn’t a lot of software in manufacturing, right? You go to any random manufacturing facility, whether it’s machining, sheet metal fabrication, PCBs, you’ll find a lot of Excel and you’ll find very little, you know, software compared to if you go to the legal space or the software space or other industries. And it’s not that, you know, anyone who’s in manufacturing can tell you, it’s not that like manufacturing people are dumb and legacy and don’t understand software.

[5:54] It’s really more that software hasn’t been built for manufacturing in the right way. And there was a few key insights that we kind of gleaned over, over my years working there. One of them was most people build software for software engineers, right? So it’s very flashy. It’s very customizable. It’s built for people who spend all day at a computer and know how to use computers really well. Turns out manufacturing, people are, you know, using their hands for building real things and they want software that just does things and isn’t, you know, doesn’t take a whole nother degree to learn how to use and learn how to manage. Yeah. I’ll give you a quote from one of my customers and I won’t, I won’t name names here, but one of my customers said the other day, if I’m going to buy the world’s most expensive software, then I don’t want to have to have my smartest people figure it out for two weeks. Yeah. So that was like the direct quote that stuck with me and it’s totally aligned with what you’re saying. You know what, what it made me think of when you were saying that was, you know, back in the mid, late nineties, early two thousands, when I was in the software business and in the manufacturing space. I mean, at the time, the world in general, like when we would implement software, like we might have to teach people how to use a mouse, right? I mean, computers that in the mid, late nineties, there was, it was very frequent that you might, you know, come across people that haven’t used one at all.

[7:12] But I remember when we started shifting to like some mobile apps, you know, there were, there were mobile apps that hit the market that just failed and no one could figure out why. And it was because like someone might be wearing gloves and they can’t push the button, right? So like we had to make buttons larger, we had to do, you know, things like that. So when you’re talking about, you know, people are working and they’re doing things and the software needs to work for them, you almost kind of have to be in the space to understand what that looks like. Absolutely. That, that, that’s a great example. And, and the second insight we learned was there’s a lot of data regulation and it’s not so simple to just give someone a piece of software, right? I mean, you, I’m sure you both have seen the news with CMMC going back and forth, suspended, not suspended. I personally think it’s, it’s going to be here to stay. We’ll, we’ll see on the timeline, but, uh, you know, there’s ITAR, there’s DFARS flowdowns, there’s CMMC, there’s CUI, there’s FCI, there’s all these acronyms. And there’s a lot of infrastructure work you have to do to be able to even give software, even if it costs 50 bucks a month or $50,000 a month, there’s a lot of infrastructure work you have to do. And so the results of those has made a very small Venn diagram where the overlap of software engineers who are, you know, want to start a company, want to build software and also understand manufacturing, uh, at a deep level is a very, very small overlap pool. And that’s why you don’t, you see like Excel is the king, right? Everyone uses Excel for everything. Now that’s not to say there’s no software.

[8:45] The, I mean, ProShop is a great example of good software. Um, but every manufacturer is going to have their, you know, CAD system, CAM system, like they’re kind of, you have to have this software to make parts. But I think maybe, maybe you’re talking more about like, you know, anything kind of beyond that for just general running your shop and handling other aspects of the business. That’s exactly right. So that, that, the last thing was, okay, we have these insights. Where do we actually build a big business? And the answer came pretty obvious to us. Uh, you know, the tagline we have is mission critical software for highly regulated manufacturing. So, uh, highly regulated means aerospace, defense, automotive, medical device, semiconductor. These are spaces where regulation, you know, it’s a good thing and make sure that they can trust these suppliers. But it also means there’s a lot of process and bureaucracy that’s required to actually be able to accept orders and ship product. Um, and I know like Mike, you probably know this at Hill. You guys, like they have lots of certifications to handle the customer base that you do. Yeah. And then the other half is mission critical. So what is mission critical? In our mind, that means, uh, essentially workflows that hit three criteria. Number one, they are a workflow that is, you know, widely done by most manufacturers, if not all in a regulated industry. Number two is there’s something that’s a daily sharp pain point. So it’s something that people deal with every single day, at least every single week.

[10:13] And then number three is something that is done today pretty manually. So if you look at those three checkboxes, that kind of rules out. We don’t want to be an ERP software. There’s already enough good ERP softwares in the market today, and they all serve good purposes. We don’t want to be another ERP software, uh, because we want to focus on like the mission critical workflows that quite literally block you from, you know, getting orders in. And shipping orders out. All right. Shop owners, listen up here. I think all of us know we are in one of the most in capital intensive businesses there is, and margins are tight and we’re, we’re always fighting issues. And then when we make money, what do we do with it? Stock markets up and down all over the place. Well, I get asked a lot about where can I invest my money? I’m a firm believer as business owners, you’ve got to invest in yourself first. You need to reinvest capital into yourself. So the number one thing that I’ve done in this industry investing was ProShop ERP. We’ve been using ProShop since 2018 and my payback literally was within weeks. With their per user based pricing, I knew when I made the decision, I only needed to save my workers roughly one hour per month.

[11:31] We saw within the first week, we were saving multiple hours per day. That type of ROI just doesn’t come along very often. Outside of that, the real value has been able to quickly implement best practices across my entire shop. From the time we get an order to the time we ship it, scaling my business with confidence, being able to go out and know that I can deliver on my promise to deliver quality parts to my customer. And then gaining insights into every job, employee, machine, and every process in our business to really drive efficiency and look for opportunities for improvement. That’s where the real value in your business drives. So visit ProShopERP.com, book a demo and start realizing ROI today. That’s ProShopERP.com. So that was kind of the core insight and okay, there is a really big business to be built here, building this mission critical software. But you need a wedge. You need that first product that’s actually going to get you, you know, your first customer to bring in revenue that’s going to help you build the team and do everything like that. And so I basically got together, joined forces with the two smartest people I knew, who were the previous, they co-founded a company with my brother, augmented reality startup, really incredible tech, deep technology. They sold it to Cadence and decided, hey, we want to take, you know, another swing.

[12:55] There’s another really, really big business to be built here. And so we got together and we, you know, we realized, oh, one obvious use case right in front of us when we went to our own factories was first article inspection reports. Right. And I’m sure Mike, you can speak on this. It may be usable, Nick, but it’s, it’s something manufacturers are doing every single day, right? They’re ballooning prints and filling inspection reports. Usually it’s done fairly manually, at least by majority. There are other softwares out there, but a lot of manufacturers are still doing it pen and paper. And it’s quite literally critical because you cannot ship your parts and get, collect revenue until you complete your FAI. Oh, that’s yeah. I mean, that’s absolutely the case. I mean, we’re, you’re right. We’re doing them every day. And, and if we do them wrong, then you can send good parts with bad paperwork and they’re rejected. Right. I mean, that’s- Describe the process at Hill for like how you’re doing it today right now, Mike. From start to finish, that might take a minute. But, um, we got a few. Yeah. I mean, a lot of it, I mean, to his point, I mean, you know, a lot of it’s, it’s not a Hill process. It’s a documented process, right? It’s a, it’s an AS9100 process or something, you know, that like we have to do it this way. So I don’t know that we’re necessarily doing it any more unique than anybody else, but traditionally, certainly very manual.

[14:19] You know, we’ve, we’ve been doing bubbled prints for quite some time now, I don’t know, seven or eight years, but it’s still, you know, we get a bubbled print and then we still have a very manual process of measuring every measurement, recording it. Um, we do it in ProShop, which helps quite a bit because there’s, they have all their first article forms. All the first article forms we’re required to provide are kind of built in, you know, and we’re using ground control and, you know, in a lot of ways it’s just, you know, it’s, it’s helping automate that process and make it where more people can be involved in that process. Right. So it doesn’t all fall on just the one person that understands the process is that, I mean, first articles sometimes for us will become our biggest bottleneck day to day, right? If we’ve got, you know, I think we were talking earlier via text. I mean, the rough math works out to, you know, I’m running 58 jobs a day. Every one of those is requiring either an internal or external first article. And you’ve only got, you know, a small handful of inspectors, right. That, that may be fully understand that process. Now, some of those are internal first articles. It’s to our criteria. Sometimes it’s to a customer’s criteria, but when that becomes your bottleneck, then all of a sudden machines, you know, if, if, if you’ve got six jobs waiting on first article, you’ve got six jobs sitting, right? Six machines, spindle not turning.

[15:53] Exactly right. And the FAI process is critical. I mean, right. For when you’re designing a new part, your, your Boeing, your SpaceX, your whomever, you want to make sure there’s some kind of process verification. Can someone build this part to spec, right? And so it’s, it’s pretty critical to get that information, make sure it’s accurate, make sure now, and then you go into, you know, then you get into PPAPs and process control and how do we build, you know, a billion iPhones, but the FAI is critical. So, okay. So that, that’s like the, you know, we had the insight, we had the team. And then the real question you get asked in Silicon Valley, but in general is why now, right? FAI has been a thing for 20, 30 plus years. Mike’s probably been seeing them for, you know, quite some time. Why is now the right time for us to start this business? Why not five years ago? Why not 20 years in the future? That’s where you see like, okay, what are like the macro trends without getting too economical? What are the macro trends here? This is a pretty fast growing market. I mean, I’m sure you guys have seen this at your own businesses, but you see the US increasing its defense budget from, you know, one to $1.5 trillion. You see SpaceX having a literally once in a generation IPO. I don’t know how many people made, you know, millions of dollars there. And those people are going to go start new companies, you know, in the next five years. You see a whole renaissance of the kind of space race, people building new satellites, people building interceptors, the golden dome project.

[17:16] There’s just so much going on this. And then there’s separately a whole trend of how do we bring manufacturing back to America, right? There’s this reshoring effort that’s going on and then credit to, you know, the US for making an attempt to do so. Cause obviously over the last 40 years, we’ve kind of lost a lot of business overseas to, you know, China and places like that. So it’s cool to see things bringing back. Well, I think a credit to the pandemic for the wake up call too. Cause I think that triggered a lot of it. Sure. But yeah, it showed you, it showed you how important it is to have, you know, domestic supply chain or at least near shore supply chain. That’s exactly right. Yeah. You need near shores, you need domestic supply chain. There’s all these new programs being launched. There’s money flying everywhere, but you know, the population of the US has not 2X. So we need to find, either we need to find people to do all these jobs and we will, or we need some level of automation, right? We definitely don’t want to take away jobs. We want to create jobs, but the math doesn’t make sense to say, you know, we’re just going to use humans to do all this extra work. You need some level of automation on the front end and the actual manufacturing process on the backend. All of those forces are coming in. And the last missing puzzle piece is, okay, how do you build really good software in AI without a team of, you know, PhDs? And that’s when you say, oh, wow, there’s this new thing called the large language model and chat GPT and Claude.

[18:42] And, you know, you were saying beforehand, Nick, like even anyone can, can write code. I would say, I think we’re going to see a lot of manufacturers building software in-house, which I think for a lot of use cases makes sense. I mean, it’s weird for me to say that as someone who is, you know, providing software solutions to manufacturing. But I think in a lot of cases, people should build things in-house. You know, your problems better than anyone else does. You can likely articulate them to, you know, Claude better than you can articulate it to some web developer somewhere. Um, now I think there are use cases where it is much more expensive and to try and build in-house than, than buy a solution off the shelf, especially when things get more complex, but I think lots of people use it. Tokens aren’t free, apparently. I’m learning this. Yeah. I get, I get put in a Claude timeout a lot. We’ve got to wait five hours. Until they reset. You know, on that, on that topic too, I think people do lose sight of, you know, building on standards and, and like good sound core data too, right? Like if I were to vibe code my own ERP system to match my operations, now I’m limited at my operations, right? Versus adopting like best practice and, you know, maybe like we learned a lot implementing ProShop, like we, we, we learned better ways to do things than we had been doing it for the previous 40 years where I’m, I fully support like using to like augment those systems. Like, I mean, we’re, we’re actively augmenting ProShop and all our different systems to aggregate data and analyze it.

[20:14] But I think people lose sight of the value of, you know, good sound, like heartbeat, right? And then from there do something with the data. Absolutely. Especially when you’re doing, and then you have like, you know, you add on the regulated ITRC UI on top of that. Now there’s a lot that goes into the infrastructure. I mean, we have a whole team that just does infrastructure to manage, you know, we see thousands of FAIs going through our platform every single month from hundreds of customers and there’s a lot of infra work to manage there that isn’t so simple, especially when you’re dealing with the, you know, regulated data. That’s the team. There’s an obvious reason to do it now. It’s a fast growing market. I mean, I’m sure you both are feeling the energy, aerospace, defense, manufacturing in America, that’s going to bleed into other end markets as well on positive kind of rising tide lifts all boats. So we, we built what we thought believed to be a best in class solution. We, you know, used our own software engineering experience, built something state of the art, built it in the cloud, wanted to keep things as modern as possible, make it as easy as possible for people, you know, at Hill or any other manufacturer to use our software without having to deal with the complexities of installations and licenses and things like that. And, and it’s just been kind of like a rocket ship ever since.

[21:33] I mean, this was what, two years ago now, we started the company. We’ve already crossed 200 customers. We’re working with mostly in the US, but customers also around the world. Small mom and pop shops all the way up to, I think like now we’re working with five public companies. And it’s just really exciting. It’s so much fun to build. It sounds like it. I mean, you can, you can hear it in your voice. You can hear the, the excitement and I agree on the boom too. Like my world’s more on the machine tool itself and all of my biggest customers are saying, Hey, we just had a record quarter. We just had a record month. So, and you know, it’s the hardware, it’s the software, it’s what’s happening, you know, in geopolitics and everything. But it seems like we, we have a really positive time and we’ve been through some seasons with making chips where everyone’s talking about recessions and how, you know, how to claw for every order you can get right now. It seems like it’s a pretty good time though. I mean, I don’t know if that’s what you’re feeling like, but it’s kind of a heyday here. Yeah. I mean, if we could get it all done, we’d have crazy growth. I mean, that’s, that’s, you know, what we’re up against is just, we’re not scaling as fast as the opportunity scaling, but we’re trying. That’s all you can do is try. And then I agree. That’s the hard thing is okay.

[22:48] I want to accept, you know, people are coming to me with orders. I’ve talked to manufacturers who are like, we’re turning down orders because can’t handle all the scale. It sounds like that’s happening for you guys as well. Yeah. I mean, that’s essentially, that’s almost where we’re at. I mean, we’re not, I wouldn’t say we’re, we’re, we’re turning down more than we have in the past, but yeah. I mean, we’re definitely in a situation where it’s, you know, until, until we fix some other stuff, we can’t take on big commitments. I think it’s like the right work too, you know, like it’s, it’s, you can’t just make the customers give you exactly what fits your shop right now. You know, it seems like where you’re, where you’re busy as heck, they want to give you even more work. And then where you need, where you need the work, it’s like, Oh, how do I find the customer who’s going to give me that type of work and it’s, it’s never as easy as it sounds when you map it all out in an Excel sheet or something like that. Right. Fellas. Hey, well, I got you both real quick. So I’m really trying to drill in like our work holding options at Hill and all of our shops. Any advice? Mike, it’s as simple as RASRAM. As what? RASRAM. You’ve never heard of it? I have not. What Nick is referring to is the seven habits of highly effective work holding. That makes more sense. Repeatability, accessibility, scalability, reliability, adaptability, and modularity. If you want to find all of these principles in one catalog, check out the SMW Autoblock catalog.

[24:13] You can find it at SMW Autoblock. E-L-O-K dot com. Thanks fellas. You had the family electronics manufacturer. That’s where the insight came from. Then you’re like, Hey, I think we could actually do this. Cause I’m not going to have to hire, like you said, 10 PhDs to get started. So the time period from like, kind of, were you working with your family’s business or were you just kind of there observing and being like, where can I create something that would help? Did you always have your mindset on like, Hey, I want to build some software for manufacturing or, or was it like, Hey, this would actually help our company and I’m doing this every day and it’s annoying, so I need to build this thing. Kind of, when did you say, Hey, like, let’s start putting pen to paper and make this? It’s a little bit of all the above, you know, we’re always ideating, looking at where we can build, where we can add value, there’s a common phrase that we wanted to avoid. It’s, it’s called, what you don’t want to do is build a solution looking for a problem. Right? What that means is like, I built some cool AI thing, widget. Now let me go find somewhere that it fits. And we wanted to avoid that at all costs. We wanted to do the opposite. We want to say, let’s find a really, really tough, hairy problem and then build the solution for it.

[25:33] Like an office space when the guy builds the jump to conclusions mat. It’s like the dumbest thing ever. Well, it sounds like you built something. Well, it doesn’t just sound like, I mean, you did build something. That’s a, that’s a pervasive problem for pretty much anyone making parts. Everyone’s got to do first article inspection. So, so that was, when you say, you know, it’s two years ago and now we’ve got, would you say a couple hundred customers now? Yeah. We, we crossed 200 customers a couple months ago. Is that really how fast it was from like deciding you’re going to have a business and starting, or did you like have the product two years ago and then get to 200? We really started this from the, from the beginning. Square one was two years ago. We wrote the first line of code in May of 2024. So two, two years and two months ago. Yeah. Good stuff. It was when we write the, and it was just, you know, me and my co-worker, we, we wrote all the software ourselves, we built everything to handle, you know, be in the government cloud, handle, at the time it was just NIST 800, you know, FedRAMP wasn’t a big thing then, and we started growing really fast. And then we came across almost an existential crisis. We started to hear people, you know, say to us, Hey, have you, have you heard about this FedRAMP thing and CMMC coming up? And, you know, we heard, we, we were aware of what’s going on, but customers started to more and more say, Hey, this might be something that’s coming up. You may want it, you know, we may need from you.

[27:03] And then it really started to pick up. I want to say September, October, 2025, we started hearing a lot of customers saying, Hey, we can’t even engage with you until you have FedRAMP because we have CMMC or, Hey, if you don’t have FedRAMP by, you know, November, 2026, when, well, again, who knows that deadline’s going to change, but at the time we can’t use your software. And that was a, you know, a pretty big deal because I would say 99% of our customers either have, or are going to have CMMC. Uh, so if we didn’t get it by November, straight, straight back to square zero. Yeah. Uh, well, and you know, I want to say, so I met you roughly a year ago, I guess, at the, um, in Indianapolis at the automated shop conference, I think at that time you were just launching, right? We probably had at that point when I met you, I want to say 70 or 80 customers. Uh, but we had, I mean, that was all just from like, you know, getting customers, guerrilla tactics, we kind of realized at that point around that July time that, oh, wow, trade shows are where you want to be. That’s where you really meet people. I don’t know if I ever would have met you and Paul outside of, you know, a random show in Indianapolis.

[28:19] Yeah. Um, and, and so that was really our like, okay, this is our go to market. Like we are going to be everywhere that our customers are. And, you know, I think 2025, we did maybe five shows this year. We’re on track to do 15 next year. We want to, you know, more than double it to over 30. Cause it’s just such a great way to meet people, shake their hand, get to know them and more than like a, you know, as much as virtual calls are nice and help you talk to anyone around the world, it’s not quite the same as if the, as if the three of us were in a room, right? Well, it’s, it’s funny. I mean, the three of us were in that room and we didn’t even meet really till like the phone call the other day. So even when you go to the same event, the chances of being able to like really connect with everyone that you could connect with, you have a limited amount of time, like you had, you had your table there or whatever I had my table there. And we didn’t connect till a year later, you know, it’s like, and I think that, you know, as far as trends and stuff, I think with all the AI and how easy it is to connect, we’ve been talking about this a lot, I think in-person community events is something that we want, we crave, like, yeah, we’re kind of all seeing each other and it’s almost like we’re in the same room cause we’ve got these webcams on and stuff, but it’s not the same. It’s not the same. You know, we could all have a plate of food in front of us, but we didn’t go to dinner together.

[29:35] You know, it’s not the same thing at all. Well, and, and, you know, let’s not discount, I mean, like let’s, let’s like detour for a second and talk about events, you know, to be even more specific, we didn’t just meet at the conference, we met when we were walking to like the event one evening, four blocks away, and we’re just happy to be walking next to each other and we’re like, yeah, what do you do? What do you do? And, you know, so like, take that, like when you go to the events, that’s great, but go do the other stuff too. You know, like, don’t just walk around, sit in sessions, go to the trade, you know, go to the tables and get free pens, like talk to people, meet them. You know, like it’s, it’s, it’s so much more than that. Yeah, you have to actively network. Don’t just show up and think that it’s going to happen. I could not agree more. That level of serendipity is like, it can’t be recreated virtually. Um, that’s why we only bring bananas, no pens. So, so people are forced to ask us what’s up with the bananas. That’s awesome. You’ll see it at IMTS. So you bring like real bananas? We bring probably, at IMTS we’ll probably go through 500, 600 bananas. That’s awesome. Every single show we bring bananas. And we’re at, my team’s at, uh, Farnborough in London right now, the air show, already gone through a hundred plus bananas, cause it’s just different. I mean, do you have like some tie-in, like, with like, oh, well the banana represents, or just, it’s just like, I don’t know, just have a banana. We’re working on something there right now.

[31:03] Look, everyone else has chocolate, candy, or like pens. And people are walking all day. They want something healthy. They want something different. It always starts a conversation. It’s never failed us. That’s awesome. That is great. Well, Hey, we’re, we’re doing all these events at IMTS where we’re doing like this happy hour and this coffee thing. And you know what? We don’t have a plan for Mike is we do not have a banana party. I think if you advertise a banana party, you might get an audience where you weren’t even ready for it. Okay. Like come to us for drinks. We’ll give you alcohol. Go to ground control. If you want something healthy. That would be the morning after the morning, the hangover cure. We’ll give you a banana. And so there you go. See, there’s the tie-in. I knew you’d see it. You need the clever founder to get to the tie-in. Well, Hey, let’s go to the name. I keep thinking about the name. Um, when did you have the aha moment? What should we call this thing that we’re building? Or did you know from the beginning? Tell us the story behind the name ground control. I wish we had known from the beginning, it would have been nice retroactively to say that. But no, we, we had no idea what to call it. We spent hours going through name generators, writing names on a whiteboard, just anything we could think of. And it’s so funny how like in the moment, everything sounds so random. And you know, you look at brands like Hill.

[32:26] Oh, Hill is a great four letter name. Easy to remember. Everything sounds so great. And then you’re trying to come up with your own name. Like, ah, this doesn’t sound right. That doesn’t sound right. And eventually, uh, we, you came up, the name generator came up with ground control. And we were like, this is actually, you can make a, uh, an argument that, you know, there’s ground control and major Tom, hopefully we don’t get sued for that. We’re not saying major Tom. There’s, you know, the actual, like, uh, Houston, we have a problem talking to, you know, the ground control, the ground station for like anything related to aerospace and defense. And it just kind of clicked and we all just looked at each other and we were like, okay, this is the one. And we bought the domain and we were just kind of off to the races and didn’t think, didn’t think about it again after that. So what were you, what was the name you were really pushing for that didn’t happen? Friends with bananas. Banana software. Uh, I don’t know if we had one. We really just, I mean, I’m curious to hear what, like, what was the, what was the story behind, uh, uh, how do you pronounce it? Goner? Yeah. Well, Hennig is a family name we acquired the company in 1999. So kind of like Mike Hill is from Cheryl Hill, a lovely lady who I got to sit next to at your daughter’s wedding and she was great, but now everyone calls them Mike Hill from Payne Manufacturing.

[33:56] Yeah. I do get called Mike Hill a lot. But yeah, no, Hennig, Hennig is like kind of the name. And then obviously Goner is the last name. We have another business called Advanced Machine and Engineering. It kind of, that’s what we do. And then, um, Slytech is kind of a play on words of Cameron Schley, the guy who created that. Um, I, I do like the manufacturing has, there’s like, you often see these words, precision, advance. It’s very easy to tell if someone’s in aerospace just from their name. Well, my 95 year old grandfather wanted a, to be starting with the letter A cause the yellow pages and, uh, we actually own AME. You get two A’s at the beginning. Yeah. Art, the Art Park Manufacturing. Um, but yeah, we have, uh, AME.com actually, like we got a three letter domain starting with A, which was wild. That just shows kind of, he was, he was thinking. That’s pretty awesome. You guys are, you guys are young enough to not remember this, but I mean, that was a real thing. Like you really were trying to be like at the top of the, in the yellow pages. I mean, but you wanted an A, a B, you know, something like that. Cause much like you don’t ever get past the first page of Google, right? You didn’t get past the first four or five lines of the yellow pages either, 40 years ago. That’s fair.

[35:20] AI is everywhere right now and it’s going to be all over IMTS, but here’s the real question. How do you turn buzzwords into actual manufacturing outcome? That’s exactly what the IMTS industrial AI conference is built for. On Wednesday, September 16th, head to the South building at IMTS for a one day practitioner focused program on reality based AI for the factory floor. This isn’t vague futurism or jargon Sue. It’s about understanding where AI really fits in manufacturing from predictive maintenance and quality to edge versus cloud, small models, LLMs, and how to move from problem to purpose, to process. If you’re trying to figure out what AI can do inside your shop, this is the room you want to be in. Learn more at the link in the show notes or head over to IMTS.com for more info. What was the minimum viable product when you first started? That’s a good question. Minimum viable product. I think for the first two months, uh, our product, you couldn’t actually rotate PDFs. So if you had a PDF that was rotated 90 degrees, you had to rotate it outside of our software and then bring it in. So it was, it was, it was very much like something that just, but it doesn’t matter, people still bought it because we realized like the core insight is you just need one magic moment that makes like is the real pain point and everything else, people will kind of, you know, over time you’ll develop the software.

[36:40] For us, that was accurately pulling text from these really old hand drawn PDFs. Like that was the thing we encountered everywhere was you see these 1960s Pratt and Whitney drawings or Rolls Royce or Lockheed or whatever. And there’s something from advanced machine and engineering that my 95 year old grandfather drew up on his drafting board. That’s exactly right. The drafting board drawings, like quite literally how they used to make drawings. And, you know, you’re trying to use, you’re trying to type this out into your AS9102 form or your FAI form, or trying to pull it out with software. And it’s, if it’s not accurate, the report will get rejected, right? Even if the parts are fine, if you have a typo, it gets rejected. So we worked really hard to build a state-of-the-art character recognition. And we built something that outperforms the rest of the market by a mile. And that was our core thing of like, Hey, you’re going to be able to pull the data accurately and quickly. And that’s true. So that like OCR, I used to have this app where I would take a picture of a business card and it would like put the text in my phone and I think the tech was called OCR. Was that, were you just like a better version of that? That’s going to like be able to interpret scribbles and stuff like that? That’s exactly right. OCR, optical character recognition, generally the default is pixel based. So it looks at the number of like literally black and white pixels in a section and says, this is most likely this, but what that doesn’t give you is some of the logic and manufacturing and, you know, you may have something that looks like an eye, but the actual image was of a dimension, like dimension that was, let’s say it’s like, you know, 1.01 inches, an eye there doesn’t make any sense, right? Anyone who looks at drawing, any human can tell you that, but if it was hand drawn, there’s some specs on the drawing, the OCR may think it’s an eye.

[38:27] And that’s the kind of mistakes we built our OCR like to avoid is how do we actually build some logic into it to say like, okay, let me pull what this is. And let me actually think about, does that make sense or not? And having that kind of two stage, two pronged approach is what let us build something that’s best in class and is able to, you know, pull from these old hand scribbled drawings with relatively high accuracy. You upload the drawing, you upload the certs, we’ll balloon everything, pull out the data and fill out, you know, an AS9102, any other FAI template. And then we will say, okay, wherever you need this data to go, we’ll give it to you in a nice clean format. It could be Excel, it could be PDF, it could be net inspect, wherever you need it to go, we can get it there. That was kind of the key is like, okay, a full product that gets you to your end state and make sure the customer is going to approve it and accept it. Cause at the end of the day, you know, from what we hear from OEMs, this does go in your scorecard. This does matter. It’s not just building good parts that are high quality. It’s also having, you know, the paperwork that lets them use those parts. So when you talk about, you know, you bubble that, send it wherever you want. I mean, one of the things you can do is send it direct to ProShop as well, right? That’s exactly right. That stemmed from that same time I met you, I met Paul Van Meter and we realized, oh, wow, we already working with quite a few ProShop customers, but for them, you know, they’re not just using this for FAIs, they’re also using it for in-process and they want these, these inspection plans in ProShop and today they’re doing it with a sort of a manual import that requires some extra work and we realized there’s a really good synergies here where we’ve already built the core ballooning technology and we have all of our kind of, you know, secret sauce there. And ProShop has built a best in class ERP that is like the source of truth of data. So why don’t we link them together since we’re both cloud applications? So we launched that and it’s been going like mad.

[40:19] I mean, I think we’re well over 40 shops who’ve signed up for this and what you’re able to do is do your ballooning, do your inspection report, click a single button and all the data transfers automatically directly into ProShop accurately as per ProShop’s kind of templates for master process control. I mean, I knew the answer to that question because we’re doing it. Um, and it’s great. I mean, it, it, it really, when I, when I was talking earlier about, you know, that bottleneck used to be in inspection, right? Like getting the bubble prints, creating the first articles, all those types of things, even just creating the forms, much less making the measurements, right? It really has opened the door now to where, I mean, hell, I could probably do it if I had to with it being integrated. I’m not going to test that theory, but yeah, no, that’s, I mean, it’s, it’s game changer for sure. Like kudos to Paul, like he’s, he gets the ProShop attention here and he doesn’t even come to the episode. It’s just, the guy just blows my mind. He’s amazing. His presence is always… Lights out podcasting. He is the perfect team. Unattended podcasting. Unattended. Lights out podcasting. No, I mean, seriously though, kudos to the whole ProShop team. They’ve been so great to work with.

[41:37] Making the integration was super seamless, getting it to customers, testing it. We got our FedRAMP certification two weeks ago, which was a huge accomplishment. I actually think we are the smallest company in the world to have FedRAMP by like number of employees, which is, you know, it’s extremely proud for us. All the work we’ve put in. ProShop just announced their FedRAMP certification, I think two days ago or maybe yesterday. Yeah, it was this week. That’s awesome. Which is huge. I mean, that’s even more huge because, you know, your ERP is where all your data lives and you really want that to be secure and, you know, make it easy for you to handle your CMFC certification. So it’s been great working with them and we’re just getting started again. There’s so much more to build. Yeah. No, jokes aside, everyone who uses ProShop seems to love it. The culture around the team there is so great. Paul’s done a really good job with that business. So you said you’re the smallest, so how many employees are you now? Right now we are 15, but we’re growing very quickly. I expect that number to increase like very, very soon. Well, I like that you said that too, because the whole idea is that, hey, if you’re in software, you’re doomed, you know, AI is going to take all your jobs and there’s no point in having people anymore because we all have Claude or whatever, and so maybe speak to that a little bit.

[42:54] Why do you need so many people if AI is a thing? Well, I would actually argue if we were pre-AI, we would probably, to support the scale we’re at, we would probably need 25, 30 people. We’ve been able to do so much more with so many fewer people. I think that’s the key, is that like, it’s not taking software. Anyone who’s used, I mean, Chachibit or Claude for more than 30 minutes knows it will still hallucinate. It’ll still mess things up. It’ll confidently tell you incorrect answers. You know, who knows what’s going to happen in 20 years from now. I can’t make any predictions on that, but today there’s still a need for a software engineer. There’s still a need for complex software that can, you know, handle data regulation, handle complex workflows, like Mike said, build best practices in given scale across hundreds of thousands of customers. I’m sure you guys are seeing it manufacturing as well, right? Like, I think there’s another saying that goes around, which is, you know, Chachibit is not going to replace you, but the person who uses Chachibit and gets a lot more done, they may replace you. So we encourage, yeah. We’re a thousand employees now. And then on my side of the business with, with me and my brother on the machine tool side, we’re like just under 300 maybe. And we, all of our leadership team has, has to go and we all want to go, but we, we have to go, we’ve hired a firm that’s like an expert in helping us use AI to like be pervasive throughout all of our systems and give us the answers quickly. We have homework assignments every week.

[44:19] We have to go to an hour long training session every week to learn how to use. In our case, it’s Cloud that we’re using, but it’s like, it’s like this whole new skill that we’re all learning at once in every, whether it’s operations or supply chain or sales and marketing or even HR, like everybody’s using it and it’s just really fun. And I love like how committed we are to it. You know, it’s not like, maybe we should use it or a couple of people can use it. It’s like, no, we’re on this mission to amplify what makes us special. Not to substitute what makes us special, but to amplify it. And so we got to use AI to do it. Yeah. I mean, it goes back to what we said earlier. We’re not, I mean, to meet the demand because of reshoring and because of explosion in the market and, you know, all the economic conditions, we’re not going to do it with bodies. The bodies don’t exist. So every body we have needs to be more productive. Right. Yeah. I mean, no different than like what you talk about on Lights Out, right? It’s not to replace people. It’s to make the people you have more productive. Yeah. Yeah. We, we can’t afford to have, uh, I call them doormen, doormen and women with, all you do is open and close the door and reload a part. That’s a robot’s job. I always talk about, Hey, the people are taking the robot’s jobs.

[45:38] You know, that’s a job for a robot. Let’s let the people do things that only people can do. And, and I think you, I see you nodding because I’m sure ground controls, like why would you have a person do that all day, every day when there’s a software that can make it way more streamlined? That’s exactly right. Every product we build, we think about that. Like, you know, one, is it the right solution? Is there a real problem here? Two, is this something where our software can help someone do 10 times? We’re not trying to take away people’s jobs, but can we help someone do when they get that surge order and, you know, Mike accepts another X million dollars of work, can his team feel prepared that they can handle all of that workload? And, and that’s where I think software in that scale really comes in handy. If you’re doing, you know, an order a day, you don’t really have that scale. I would actually say you don’t need software. It’s not, it’s really for people, you know, when you’re trying to scale, when you’re in that growth mode is where it makes a lot of sense and that’s what we think about, you know, we won’t have time to get into all the other products, but as a teaser that the new product we launched, which we’re really excited about is, um, contractor view, you know, anyone who’s reviewed contracts from the primes, from the Neo primes, uh, you know, they know they can get pretty hairy. There’s terms and conditions. They switched the unit price from the quote, which on a big order can make quite a big difference. And we built a really cool system that actually will take all the ingested data, the emails, the quote, the PO, and tell you very quickly, Hey, here’s the standard checklist you should be looking at, here’s what they messed up. Here’s cited sources of where in each document we found that they made a mistake, and then here’s an auto-drafted email you can send to the buyer, telling them where they messed up and how to fix the PO so you can accept the work faster.

[47:21] So I didn’t know you’ve done that. That’s cool. I want to learn more about that and I’ll give you the reason why. So Nick knows this, I’ve talked about it, you know, I’ve over the past month or so, something like that, I’ve asked my entire management team to identify like these big time sinks that they’re dealing with, right? And where are you looking for information that if you already knew it would, you know, you can make a decision in two minutes, but you’re spending two hours getting the information you need. And in a lot of ways to figure out like, where can AI help us? Taking a lot of those answers, feeding them to AI and saying, Hey, look at, look at this and tell me where you can help me. And it’s funny, you talk about the contract review because what it identified that I would have totally skipped over is almost every response I got came back to the PO entry process, could create these major time sinks. One, entering the PO takes time. Doing contract review takes time. But if, you know, if there’s anything wrong on it, we might make the wrong part of the wrong rev love, right? And we got to start over or, you know, you get all the way at the end and your invoice and your PO was in wrong, like your invoice might get rejected, right? So that PO entry process and contract review kind of came back as this like weird bottleneck that we didn’t even know we had across our entire value chain.

[48:44] Right. Like from essentially from quoting all the way to invoicing. And so that’s actually something that we’re trying to tackle internally right now is like, how do we improve that PO entry and contract review process to eliminate all these potential errors? And that was the thing is the errors that can come out of that process. And they’re rather, they’re fairly infrequent, but they’re incredibly costly. Right? So, yeah, exactly. It’s like, that’s like one of those cases similar to quality where you don’t want to just chase time savings, right? You want to chase time savings, but really you want to chase accuracy. Cause if you have a six, seven figure PO and you make a mistake, you know, I don’t think every customer is going to say, Oh, no worries, let’s rewind everything. Let’s take care of that. I’m going to brag on one of your products that I just love that we use. I don’t even know what you call it. The check-in system. The check-in system, you know, like you, so when we have visitors, you know, we’re ITAR, we got to, you know, register everybody. There’s a QR code on our, when you walk into our lobby, you scan that, you fill your thing out, it notifies whoever they’re there to see that they’re there at the lobby, like it’s great. And it’s such a, such a quick, simple little solution that is incredibly valuable, saves time and honestly, just kind of looks better. Right.

[50:13] Than the 20 year old pads sitting on the desk that everybody’s got to walk up to and fill out. I know. I’m glad you brought that up. It makes it feel like you’re a certain type of shop when you got, when you have like a, a kind of an iPad process and it prints out the little sticker or whatever, you know, it’s looks nice. So. That, that, that’s exactly what it is. Like, Hey, this is a, you go to any shop and you see the paper everywhere. And then you tell someone, Hey, if someone audits you, you need to have five years of logs and prove citizenship. That’s a tough ask. And so our thing was like, build something easy, implement it in five minutes. It just works. You don’t need to do any customization or set up, but you can, if you want to, but you don’t need to, and that, that’s kind of our whole thesis at Ground Control. So Nick, Nick, sometimes we send that link to people that we know are coming. So they get it done before. Yeah. Wow. Yeah. So they’re, I mean, basically like they’ve already filled and, and you do it from your phone, right? So all your auto fills already there. It takes you four and a half seconds to fill out the form. That’s cool. There was a event, Pro Shop event where they hosted it at a shop that uses our check-in system. And it was great because they had, I think, 30, 40 people coming for the event. Everyone can just check on parallel as they come. You’re not waiting in line and iPad. On your contract review stuff, does it review NDAs and stuff like that too, or? So the core use case really is POs and quotes because we want, like, we didn’t want to just become a competitor to ChatGPT, suffice to say, you know, we’re going to keep building on our product roadmap and, and we have heard this from, from multiple customers, what you just asked me, so I’ll say. Sometimes you get like the basic boilerplate one and it’s like, okay, this is fine and then other times they give you like a Bible and you’re like, well, what, like, what can I not do anymore now? You know, like you basically, you get so paranoid by the time you’re done reading it. You’re like, are we allowed to make anything ever again? If I sign this? Yeah, that’s pretty fair.

[52:05] Yeah. Yeah. Cool. Well, I’m sure there will be many updates to the Ground Control story and thanks for kind of giving us the origins and, and where you’re at today as we close here, give us kind of the one thing you’re most excited about and the one thing you hope to have a lot of conversations about at like IMTS, for example. Great question. Well, first it was, it was great for, for being here. Thank you both for having me. We’ll definitely see you both at IMTS. I think one big thing we’re definitely excited about is having this FedRAMP certification because it really gives us a way to bring all of our technology, current and future products in a compliant way and make sure that, you know, our customers don’t have to worry about, doesn’t meet this standard, doesn’t meet that standard. It just works and you can use it. It’s compliant. That’s what we’re really excited about because that unlocks a whole new wave of growth, right? That unlocks for us, a lot of growth and it unlocks for our customers, a lot of growth and a lot of, you know, not having to worry about these things. And then number two, I think we’ll be having a lot of conversations about our contracts review product as we bring it to market and understanding, you know, this is such a general process. There’s not really a software that I know of that does what we’re building. Okay. You know, our ballooning software, there are other FAI generation. There are other products in the market, but for this, there’s not really anything that exists today. And so it’s going to be really interesting to see, can we carve a new niche in this market? Can we really solve this properly across, you know, 10 different manufacturers may have 10 different checklists.

[53:30] So how do we really build that source of truth and make it generalizable? Yeah. Niche is the word for me because like there’s legal AIs that’ll help you review legal stuff. And you’re probably paying for a bunch of stuff that you don’t actually need on a day-to-day basis. The way you would need like a manufacturing specific solution that’s, you know, a shop like Hill or Hennig would want to use. So I’m excited to see where you guys are going with it and congratulations on all the success so far. We love a good entrepreneur. Don’t we, Mike? Love it. Yeah. Awesome. Thank you both for having me. Yeah. Well, I think you said it earlier, you know, if you’re, if you’re waiting for six hours for a bunch of first articles to happen, I think it was you who said it, Mike, you know, if you’re not making chips, you’re not making money. Bam. If you’re not making chips, you’re not making money, you’re not making money. Bam.

[54:47] If you’re not making chips, you’re not making money, you’re not making money. Bam. If you’re not making chips, you’re not making money, you’re not making money. Ladies and gentlemen, if you’re not making chips, you’re not making money, money, money, money.

This transcript was auto-generated and may contain errors.

Connect with Mehul Shah

Connect with MakingChips

Subscribe

For the latest episode straight to your inbox, business-building tips and other exclusive content, Subscribe to #MetalworkingNation today.