A digitized solution to managing manufacturing workflow processes w/ Praxie CEO, Michael Lynch

We like to talk to interesting people on this show so when Praxie reached out to us wondering if their CEO, Michael Lynch could come on the show we knew it would be a fun conversation. 

Michael is the former head of Internet of Things at SAP software and now CEO of Praxie (and believe it or not, former lead in Les Miserables on Broadway prior to that!). 
We talk about how Michael went from the acting/performance world to the internet of things and eventually wound up founding Praxie. 

Praxie provides a platform that allows organizations to transform their workflows with AI-powered processes, dashboards, and applications, without spending an arm and a leg and taking months to complete. This is especially relevant for manufacturing businesses who are stuck using Excel spreadsheets and software from the 80s. 

pickplacepodcast.com

Transcript

Chris: 0:08

Welcome to the PickPlace podcast, a show where we talk about electronics manufacturing and everything related to getting a circuit board into the world. This is Chris Denny with Worthington.

Melissa: 0:17

And this is Melissa Hough with CircuitHub.

Chris: 0:19

Do you think people aren't quite sure whose voice is who at this point? Do we keep needing to tell people who we are at the intro?

Melissa: 0:26

Well, for the new people, the new listeners, they won't know.

Chris: 0:28

that's true. That's true. That's a good point. That's a good point. All right. Solid point. We'll keep doing it then. What's new, Melissa?

Melissa: 0:35

It's been a

Chris: 0:36

cat this

Melissa: 0:36

return not in this episode. Oh, I don't know. I feel like things are always super busy towards the end of the year, work wise. Everyone needs something. I'm sure, yeah. I know it's been crazy busy at the factory, too.

Chris: 0:51

crazy busy. We have not been this busy in years, honestly, pre, pre COVID. We haven't been this busy since pre COVID, I think. And it, but we're executing, it's brilliant. We had time to kind of breathe a little bit and set ourselves up for success and we're just pulling it off. It's really exciting. But on both sides, both on CircuitHub's side and on Worthington's side. Our guess is that it's like, a lot of people want to wrap projects up by the end of the year. They, they have budgets they want to spend, that kind of stuff, yeah, so we get super busy, but, yeah. So it's so easy to ignore, you know, recording the episode. Ah, just push it off another week. We'll push it off another week. But, uh, this week we're, we're excited to have a special guest. We're going to have Michael Lynch on the show. He is the former head of the internet of things at SAP software. Maybe you've heard of them. And now the CEO of Praxie and believe it or not, and I can't wait to talk about this. The former lead. in Les Miserables on Broadway. I have got to know more about that. Michael, welcome to the show.

Michael: 1:52

Hey guys, hi Chris, hi Melissa, how are you guys today?

Melissa: 1:55

Good. How are you?

Michael: 1:56

Excellent.

Chris: 1:57

Excellent. Me too. Me too. All right. All right. Dish. Cause you, like my wife and I have seen Les Mis on Broadway at least four or five times now. We've seen it in Boston, but like it is, I don't think we're alone in saying this, our favorite Broadway show we've ever seen. And we've seen dozens of them.

Michael: 2:14

It's pretty good. There are some other ones, there's some good stuff in like, you know, Wicked,

Chris: 2:20

We just like being really

Michael: 2:21

Hamilton.

Chris: 2:22

like being sad is

Michael: 2:23

You want to bawl your face out? Yeah, I dream the dream, yes. So, the story, so I actually, yeah, unlike you, Smarty Pants, I I went to music school at the

Chris: 2:33

Oh, cool.

Michael: 2:33

School of Music. And and then I actually dropped out because it was, as you do, because I, I realized I was spending all this money and nobody cared about, in, in theater, nobody cares about your degree. They just care, you know, can you sing? Can you dance? Can you act? What do you look like? And so I went, I've kind of created my own curriculum and just studied for a few years in Manhattan. And then I started getting jobs and ended up doing a bunch of Broadway shows. And then I was on general hospital for a couple of years.

Chris: 3:01

Come on.

Michael: 3:02

Yeah, trying to be a heartthrob.

Chris: 3:04

Like out in Hollywood.

Michael: 3:05

yeah, yeah, yeah, yeah. My wife and I both came out. She was on Seinfeld and a whole bunch of other things. She's done a ton of stuff. She's done a ton on Broadway as well.

Chris: 3:14

Is that how you get met? You met in the arts?

Michael: 3:16

Actually, you know what's really the coolest thing is we have a video of the moment we met.

Chris: 3:20

Come on.

Melissa: 3:21

Aw.

Michael: 3:22

because we were doing a TV commercial and we were dancing on it and I came in and she came in and we have the video of us rehearsing. And so that was, but she was dating this other guy and I would see them all around, you know, they, we,

Chris: 3:34

him.

Michael: 3:35

well, he was like, I wasn't like, it was weird. It was like, we knew each other and I'd say, Hey, how's Robbie? And then yours went by and how's Robbie? She goes, Oh, I broke up with him. I got, Oh, that's terrible. What are you doing? So, so, that's how, that's how it all came together.

Chris: 3:51

too cool.

Michael: 3:52

yeah, so we've been together decades now. It's great.

Chris: 3:55

Nice. Alright, so, no, like, what was your role on Les Mis?

Michael: 3:59

Oh, so I played Marius. I did the empty chairs at empty tables.

Chris: 4:03

No

Michael: 4:04

Yeah, that's the song I did. Yeah. Yeah. And I'm the, I'm the one that lived,

Chris: 4:09

See, yeah, my wife, I think that's my wife's favorite character. I'm trying to remember. I'm

Michael: 4:13

my wife had the better character. My wife played Eponine.

Chris: 4:16

Oh my gosh, you're kidding

Michael: 4:18

Yeah. So we did it together and she would dime arms every night. And the fun, the story, the funny story I tell is the

Chris: 4:23

And you were dating at the time, or were

Michael: 4:25

we were married, no, we were married, they hired us as a package. So,

Chris: 4:28

This is incredible!

Michael: 4:30

it was cool. She would, she would dye my arms into a win-win rehearsal, like. Well, you didn't have to act. I was like, well, she's all covered with blood and crying and then by about the fourth week You're like your elbows in like stop it, you know So, yeah, we did that we did it for about a year together

Chris: 4:50

That's so exciting!

Michael: 4:51

Came out to Hollywood and did some stuff there

Chris: 4:53

We talk, we talk in like, late 90s, 2000s, like, what,

Michael: 4:57

90s yeah 90s and then I came out to LA and I was doing soap operas and stuff and I I actually, you guys will laugh because you're younger than I am, but I have this Time Magazine cover of what they were, they were talking about the internet.

Chris: 5:08

Oh,

Michael: 5:09

And I was, yeah, I was sitting in my, I was sitting in my dressing room on Broadway looking at the same going on. It was like that, that movie where they say plastics, like I got to get out of here.

Chris: 5:18

Yeah, the graduate.

Michael: 5:20

the graduate, right? Plastics. So I was like the internet. So I went to Hollywood and I was doing TV shows and I kept looking for At that point, I had got the bug. I had to get it and I was doing well in acting and everything, but I just didn't feel like I didn't really want to be an actor. I kind of wanted to be a pop star and all that stuff. And I, I was losing my hair at that point. I go, that's over. So I got to do something cool. So anyway, I got into the internet. I got, I walked into a, yeah, I walked into a computer game company. That was a startup that I had read about and said, I'd work for free. And I go, Oh, free is good. And so I came in and and then, I started building stuff and I ended up running production and we took the company public. So I got the sort of on the ground. I never, I mean, my background is just weird. That, that was really lucky. I happened to walk into a great place and we figured it out. And so I learned that whole process. It was called Seventh Level. They eventually, we went public and then they sold it. I can't remember who sold it to in the end.

Chris: 6:15

Were they known for any particular Did they have any particular

Michael: 6:18

This was early 90s. So you guys probably, do you remember all these games where you had walked through hallways and try to figure out the combination and something would open up like Myst? Yeah, we did. The most famous one we did was one based on the Monty Python series.

Chris: 6:31

Oh, no sir.

Michael: 6:32

yeah, so what the best job I ever had I spent three weeks transcribing Monty Python So I got paid to sit there and watch Monty Python all day long and laugh and change, you know

Chris: 6:44

That's incredible.

Michael: 6:45

that is an ex parrot, you

Chris: 6:47

That's incredible. That must have been amazing.

Michael: 6:50

it was great.

Chris: 6:50

I, I, this show is known for going off in tangents. I'll just share a funny story about transcribing things. My sister was doing something in college. And she told me, she's like, I'll give you 40 bucks if you transcribe this, this thing. thing for my paper. And I'm like, cause I'm pretty fast typist. I'm like, sure. And so she's interviewing this woman. She, she went to school to be a dietician. She's interviewing this woman and this woman is explaining how after she had a child, she's having trouble, like losing weight again, she wants to, you know, feel comfortable again, yada, yada, yada. And, and my sister's like, Oh, I know. My mom told me that if she had known how much weight she was going to gain when she had my brother, she never would have done it. I'm like, Hold on a second! I'm sitting here transcribing

Michael: 7:30

never told me that.

Chris: 7:33

So careful when you're transcribing things. You may hear things you never heard before.

Michael: 7:36

no, it was super fun. That sounds like, that sounds almost as fun. A little, little harder, but almost as fun.

Chris: 7:41

Yeah

Michael: 7:42

and then started making computer games. Then the 3D aspects got me into mechanical cad, and then we created a company that read hundreds of different mechanical CAD formats and we generated illustrations and virtual reality and all that kinda stuff. And then SAP bought that company. And we're basically their collaborative backbone and visualization backbone on the SAP platform. And then I, and then I took on the Internet of Things and Industry 4. 0 for them as

Chris: 8:06

What time frame, roughly, like when internet 4. 0 and

Michael: 8:10

So they bought our company in 2012, 13. And so, you know, that decade you know, the Internet of Things and, you know, magic machines and all that stuff was kind of a super hype thing. And as I. You know, got and tried to communicate that, you know, manufacturing had been doing it for a while and they were, the question was, could you take it much more broadly? Like everybody tried to put it in coffee machines. I mean, I had, I had dozens and dozens of meetings at SAP with vendors who were trying to figure out how to get it into machines that didn't really need it.

Chris: 8:42

Well, it's here now. Like, my dishwasher is connected to Wi Fi. Like,

Michael: 8:46

Yeah, it's here. Yeah. And the vision of it was, it was like all encompassing, but I think, I think the internet is like You know, like motors where they thought, I don't know if you know this history. They, they, when motors, electric motors first came out, they thought the house would have a central electric motor.

Chris: 9:02

Yeah, because that's how factories were,

Michael: 9:04

Yeah. And, and then we ended up with motors and everything. And so I think that's the truth about internet as well as, you know, AI and everything else that comes along. It'll, it'll just micro into the app and it'll do what it's supposed to do.

Chris: 9:14

Yeah, yeah, yeah. Fascinating. All right. So you're at SAP.

Michael: 9:20

Yeah,

Chris: 9:20

I don't know, you get bored or whatever. What, tell us about Praxie and, and how, how you got started with that? Well, first of all, I think I, I did mention it yet. You're now the CEO of Praxi. Tell us a little bit about like what exactly it is and, and what you're trying to do with it. When, when did you start it?

Michael: 9:36

It's about five years old now. We. So this goes to SAP. SAP is an incredible company and you know, the amount of infrastructure they have at manufacturers around the world and I was, we sold our product into the manufacturing group and then I was right in the dead center of them trying to figure out how to take this to a platform level instead of the specific applications like your ERP and your, your MES system, et cetera. But every manufacturing plan I went into, I saw core applications, whether they be SAP or not. And then I saw Excel everywhere.

Chris: 10:09

Yes.

Michael: 10:10

thought, why is that? And, and you know, the more I learned about it, the reason is because every manufacturing company and even probably two PCB manufacturers do it differently. And

Chris: 10:21

Yes, of course we do.

Michael: 10:22

yeah. And Excel is an incredibly useful and valuable way to hack an app together to do something

Chris: 10:29

That's right. Yeah. And we, we basically do that, but with Google Sheets, right. And

Michael: 10:33

same thing, Excel. I mean, it's a,

Chris: 10:35

and you, you can, it's such a general purpose tool.

Michael: 10:40

it's a great tool, probably the best office tool ever made. buT. It has real limitations. Everything has to be a grid. Every rolling up data across lots of different sheets is really problematic. Getting people to know where the sheets are is problematic. You end up in a, you end up in this conundrum that a best practice is not always best laid out in a grid and a way to manage is not always best. So what we did, what Praxie is, Praxie means the practical application of knowledge. And what we were trying to do is create best practices that could be. Instantly modified for customers. We, we, we stated that we wanted to make enterprise software as easy as making PowerPoint.

Chris: 11:18

Okay.

Michael: 11:19

And we're not there.

Chris: 11:20

It sounds like HyperCard, because I always think of like HyperCard

Michael: 11:24

Dude, that is so fun. It's, it's, it's the, the internet's version of HyperCard with, got, getting rid of a lot of the problems that HyperCard has, that's what we built. And basically this infrastructure is unique. You don't build applications by building backend structures and your relational databases, and then building your business logic and then building your UIs. You basically just make what you want on the screen and then you extrapolate out the backend and it will just functionally create. The structures on the back end as you structure it. So we can build full applications in days. And they're very stable because it's all based on a no code code base. And they're very scalable because of the way we built the back end. It's as scalable as, you know, Google and it's on their infrastructure, in fact. So what we are trying to do. Is basically say, you know, in a world where you need, where AI is going to be important, everything has to be digitized and it has to be digitized in a scalable way where the data is auditable and all those kinds of things. And if you do that, you can move at web speed on all kinds of applications. And we want to remove the barriers and those barriers. My experience was barriers. Okay. At SAP, we go, Oh, you want this new scheduling system. We can build you that, you know, and guys in thousand dollar suits had come in and negotiate a deal and. A few million dollars later, we would have a plan to build something that would take 18 months. What we want, what we wanted, and that's, and it's incredibly good. Don't get me wrong, I'm not bashing SAP. Incredibly good. But what we thought is there's lots of companies that would never be able to afford that. They can't, they don't have, their IT teams are not big enough to build that kind of stuff. They can't take, they don't have the expertise to take other applications and try to extend them. So we want to be able to come in and deliver an application to do that for. You know, the kind of money that they get the return on in three to four months and get them up and running in four weeks.

Chris: 13:15

So, let me, let me just hit pause there for a second. I, cause I have, I have a number of questions based on what you just said. Melissa's rolling her eyes right now. Like, okay, here we go. So, It sounds to me, I feel like, and apologies if I'm bringing this up and you view them as a direct competitor, but it sounds to me a little bit like a software I'm familiar with called Retool.

Melissa: 13:41

I was literally thinking the exact same thing.

Chris: 13:43

sounds familiar and I don't know if you can consider them a competitor or not. Is this like a new class of software where, yeah, now like we're realizing Excel is amazing, but we, we need, you know, we need proper databases, not just a big table. Like, and we need, we need forms and we need table views and stuff like that. And am I on the right track?

Michael: 14:07

there. Yeah. And the, there is a mountain of peak companies that, that are no code fast software developers

Chris: 14:15

Okay. Okay.

Michael: 14:23

to create a better way to code.

Chris: 14:27

Interesting.

Michael: 14:28

We come in with these frameworks. We have the entire Lean methodologies framework. So we come in with a full AI enabled GimbaWalk and say, how do you want to run your GimbaWalk? And then we'll turn that into your GimbaWalk in days.

Chris: 14:42

Hmm.

Michael: 14:43

So we're not, so we're, we're not trying to code better. We're trying to deliver best practice workflows and processes faster. Now they'll end up in the same place on some level, but one is a bag of Legos and one is a best practice. 90 percent there that will turn around very rapidly

Chris: 14:58

Sure. Yeah. Yeah. Yeah. Well, for sure. I, you know. There's, there's different approaches to trying to you know, accomplish the same thing. I, I guess what I'm curious about is with a tool like Praxie do you find that people are generally like, like signing up and then going, okay, Michael, I've got it. I'm going to use your tool to do, to build my thing. Or are they coming to you and saying, hey, Michael, like here, here's what we need. Here's what we need. Can you help us build it?

Michael: 15:26

I would say 50 50. Yeah. I think it's more of this is, I don't like saying this and this gets maybe into pet peeves, but, but. I think that I've never been at a manufacturing company where there are extra people.

Chris: 15:44

No, they don't exist. No.

Michael: 15:47

Right.

Chris: 15:48

No, we run negative most of

Michael: 15:49

right, exactly. So we built Praxie so that any knucklehead like me could build enterprise software. It's really easy when you think of the difficulty of building enterprise software. It's about as hard as mid level Excel.

Chris: 16:03

Yeah.

Michael: 16:04

Like you can build, I build crazy stuff in days. I built a whole CRM system in a weekend.

Chris: 16:09

Brilliant. Brilliant.

Michael: 16:11

But, to learn how to build in Praxy, requires somebody learn how to build in Praxy. And my experience is nobody has time to learn a new tool. So a lot of customers go, that's great. We know we can, and here's the, the, the general process we go through. We'll come in and somebody will say, we need to digitize our daily, you know, end of shift reporting and SQDC boards. We need, we need that digital. Cause we don't have any visibility. We can't tell where the

Chris: 16:36

alphabet soup I have not heard before. And I've heard a lot of alphabet

Michael: 16:40

Oh, you haven't heard this one?

Chris: 16:41

me now.

Michael: 16:51

end of shift reporting is a really simple one. If you don't have the data coming off your, you know, your, your machines and dynamically building it, you generally, there's most small companies, there's a person who at the end of the day says how many were made, how many were scrapped, et cetera. So they want to digitize all that. So you roll up into a dashboard. Everybody can see how the day went. Did you win the day? Did you win the month? We'll start on something really simple like that. And we'll build that for somebody. And then they go, wow, we need this problem and this problem and this problem. And pretty soon we're seven or eight applications in, at which point they go, we want to have, we want to have a dedicated person on our team, really learn how to build everything and just build everything for us.

Chris: 17:25

Yep.

Michael: 17:26

So we, that's generally how it rolls out.

Chris: 17:28

makes sense. That makes

Michael: 17:29

And, and we do, because we can do it really fast, we can do it cheap. Like the other thing we try to do is a 10th of the time, a 10th of the cost.

Chris: 17:35

Yeah, well, that would that be nice. That's a winning formula.

Michael: 17:39

And we can do it because we, you know, add SAP, like, you know, it's millions of dollars and any, any big enterprise software, it's very expensive. It's time consuming. It's rock solid, but it takes a long time. In our case, we can, you know, take on that, you know, like your company with 40 or 80 people between you, I don't know, you know, we can, we can turn all your Excel sheets into applications at a price you guys would go, wow, that's awesome.

Chris: 18:00

So let me let me. This is this is actually the timing on this is is very interesting to me because we have for many, many years and I may have. We've talked about this on the show before. I can't remember everything we've discussed, but, For many years, we've been operating on an MRP software called Alliance. Now Alliance was written, I'm pretty sure in the 80s. I'm fairly confident it was written in the 80s. It's database backend

Michael: 18:26

Hey, man, don't, don't beef the 80s.

Chris: 18:33

specifying it that way. The, the backend database is old school Microsoft Access. And it. only runs on a 32 bit operating system. You cannot run it on a modern operating system. So we have all these virtual machines that virtualize old, like Windows 10, Windows 10, but 32 bit so that we can get the software to run. So,

Michael: 18:55

thinking you're running on a virtual Nintendo or

Chris: 18:58

yeah, don't tell anybody I got virtual Nintendo in my, in my office. So, it works though. Right. It works. There's no support. I mean, the, the, the manufacturer of the software has given up support on this decades ago. Literally probably over 20 years ago, they no longer supported it. And they have modern versions of it. That, that are very nice. We own it. It's free and it works. And it's one of those things where you see all the time, like I go to like my local tire store and they've got like an. Old DOS terminal in a virtual machine that they run their entire operation on. You see it all the time. They're running this old software, but we know, we know in our heart of hearts that that is not, you know, we're going to maximize that tool, but it's not going to help us grow, right? And we want to grow. We're, it's not going to take us to the next level. It's, it's sort of like when we switched from well, it. Very, very specifically, you may not be familiar with it, but in the circuit board assembly, you have machines called pick and place machines. And the platform we were on previously was a fine tool, got the job done. We were delivering boards, customers were happy, but we knew it wasn't the platform that was going to help us to grow. And we knew, even though we own those machines, we knew we had to spend the hundreds of thousands of dollars and move on to a platform that was going to help us to grow. And sure enough, we've made that transition. It's been a little over a year now. And we're not looking back. It's just, it's been a fantastic transition. We're very excited about it. So we're in this sort of state with our current business software, right? This, this old school, it's called Alliance software. We're, we know we have to move off of it. We know we want to find a tool that allows us to streamline our processes. And I, that is so buzzwordy. I hate, I hate saying it. It's so buzzwordy, but it's like, it's like literally there is, you know. Like, I'll, I'll describe for our audience here, the listeners, we, we, we get your design files, right? And when we open up those design files, we identify, let's say, for example, the dimension of the, of the board, okay? So now we have the dimension of the board. But then the stencil guy has to get the dimension of the board. The pick and place program guy has to get the dimension of the board. The, the stencil program, the SPI program, the AOI program, the reflow oven, the selective side, we all need. The dimension of the board, and we all look it up each and every time we need it. Right? So we're okay. It's 121 millimeters by 160 millimeters every single time. And you have that times 10 different processes, right? Whereas a modern platform, you would be able to plug in the dimension of the board, 121, 160, blah, blah, blah. And then everybody, you would

Michael: 21:48

propagates to everybody who

Chris: 21:49

Just boom. That's what I mean by streamlined processes. It just goes everywhere all at once, rather than having to redundantly enter this information over and over. Our current tools are not capable of that. And those are the things we miss out on when we don't move to a new platform, which it sounds like Praxie is the kind of a platform that we could move to that would help us to do this. Now, the big question is for a guy like me is. It's, it's not free. I'm not asking you to give your pricing. It's not free, but you have to make a certain level of investment. And every time we've talked to vendors of classic ERP, MRP software It's way too much. I mean, it's not just Mike, Michael, it's not too much. It's way too much. Like be like, we would have to be literally 10 times the size we are today to have the scale, to be able to afford what they're trying to charge for the service.

Michael: 22:49

And that's the conundrum. Right. The alternative is you have your IT team build a custom solution for you.

Chris: 22:56

Which is, we're almost doing it because we need these tools and we've started to do it, you

Michael: 23:01

And you think it's cheaper because Fred or Sarah is already on the team.

Chris: 23:06

the reality

Michael: 23:07

the reality is what you're doing is you're locking in a very slow development and it's going to add, you're going to add two additional people and the system is going to become very brittle. Right. Because they're not core, they're not a software company. They're just two people with great skills.

Chris: 23:21

going to be very brittle because they're not for, they're not a software company with great skills.

Michael: 23:30

yeah.

Chris: 23:33

Google hires them, you know what I mean? Like, oh, and now we're stuck. It doesn't matter how good his code is, how, how well he commented everything. To get somebody else up to speed is going to

Michael: 23:43

Even if none of that happens, even if none of that happens, it'll be a lot more expensive than you expect. I was on the phone the other day with a group who had hired a guy for 25, 000 to build a tool they needed just to move their, they do metal CNC ing, just to move their work orders across the process. It was pretty simple process. They were 250, 000 in

Chris: 24:05

Pooh. Oh, yeah.

Michael: 24:09

it. And so, and that's, you know, to a lot of companies, that's nothing, but to smaller companies, you know, 25, 50, 000, this is real money and it has to work and it has to be scalable and all that kind of stuff. So, one of the things that we often do though, is we don't try to replace core MES or ERP system if it's, if you're managing your inventory and your, you know, your materials and all those kinds of things well. We don't try to replace that. What we tend to do, like your system, your Excel system, or your access system, is if there's a way we can talk to that. And I don't know, use the data that's, yeah,

Chris: 24:42

so, I'm so sick of supporting it. I'm so sick of fighting with virtual machines and trying to get the virtual machines to print, to act, you know, real world printers. No, I'm done. I'm done.

Michael: 24:54

it, so if you like, or you want to keep certain software, we tend to talk to it and what, and. And we can go into in depth at whatever time you want, but basically the way our system works is think of a giant data lake that surrounds your existing applications, your existing databases, and your machine data. And that data lake can bring in any of that information, and then we can use it to build and work with on any application that you want. So instead of having, you know, lots of various applications and Excel sheets and connectors and ad hoc workflows, you have basically a unified structure to bring everything together. And then on that structure, you can capitalize on to build data, but you can also connect data that you would never would. You'd be like, okay, well, how many of these, you know, PCBs are going from here to there and actually getting through the process. Cause right now they're on two different apps or two different Excel sheets. And we have no ability to understand the flow. So once we get them on a data structure, we can then create reports that will show you how this connects to that, et cetera, et cetera. So one customer right now,

Chris: 25:53

I was just going to say, we've been talking about the forest. Let's talk about a tree. Can we, can we point to a specific example, like something you don't have to say, Oh, it was Walmart or

Michael: 26:01

no, no, sir. Sure.

Chris: 26:02

you know.

Michael: 26:02

So one of the things that I find is in manufacturing companies is they, the, I don't know, there was an old politician who said the unknown unknowns.

Chris: 26:14

yeah, yeah. Was it Rumsfeld? I

Michael: 26:15

I think it was Rumsfeld, right? The Unknown Unknowns. So,

Chris: 26:19

knowns, known unknowns and unknown

Michael: 26:21

Unknown Unknowns, right? So it's the Unknown Unknowns that really get you. And, and so what we tend to do is try to get, digitize the plant so that the core information is coming up. Into shift information, all the lean stuff, you know, you can input cool processes like 5S all the stuff for Incident management, maintenance, all those kinds of things. So you get that stuff digital so you can see the stuff, and then you start seeing those unknown unknowns. You start seeing that some, why do we have so many people cutting their leg? Well, because, you know, whatever it is. My point is often,

Chris: 26:56

it's We're running out

Michael: 26:58

yeah, whatever it is, but that's a really kind of simple one, but the idea.

Chris: 27:02

the cutting the leg problem. Just get more band aids. We don't get

Michael: 27:06

there's a lean concept called the Obeya room command center kind of a thing. So the basic idea is get the core processes digitized in a way that you can see the core information that drives OEE or whatever calculation model you're trying to drive. In the things that matter on your verticals, that's the safety, quality, et cetera. First time quality, first pass yield, all the different things that you're trying to get. Get the data coming in so you can see it and everybody can look at it. And then we go after what is the infrastructure required to optimize that further. So we need to connect to your ERP system. So I'll give you a couple of examples. Working with a client right now who they can't ship their products because testing. Is a problem. They, they, so the way they have discrete kind of batch products. And they, they test each batch product before they send it out. So, a sales person or an ops person says, this is the group of stuff we need to test. They send that in an email. The testing engineers start testing stuff. Then they get another email. And somebody's higher up on the food chain says, No, test this product right now. A customer sent it in. I need to understand why it's broken. So they put down the group they were working on and start working on another product, etc. So they can't see the visibility into their priority list. They can't see who requested it. They can't see which orders it's affecting. So you build a system that now they have a work order system for their engineering test. Fantastic. So then the engineers are also looking at program management. And then they want to be able to have a handover that sends, one, to identify which programs they should be doing, but then at the end, when they're ready to send a program over to an engineering test, connects the dots to that. Oh, and then the suppliers are sending in Products that relate to those overall products that need to be tested, and they need to be able to run supplier testing, and they want that to connect to the testing that's going on in the engineer. So each one of these are steps in getting the product out the door. That are currently being managed in Excel sheets and emails that give them no visibility to where the hangup is in that test product process. So by connecting that data structure, they now, and creating workflows that everybody can see, and people are notified via email and all that kind of stuff. What you're, you know, it's just process improvement. You can now see the problem. Why does it take us six months to release a product? You know, I can't, I don't understand why you can now see the problem and get it digitized. The problem is that, and that's a really basic sort of.

Chris: 29:29

just an

Michael: 29:30

Non lean problem. It's just an operational workflow problem. But I, what I see this all the time where people are running Excel and they don't know what they don't know and they don't know how inefficient their process is.

Chris: 29:42

Not, not my process. I have an efficient

Michael: 29:44

No, yours is great. Yours is great. by the way, you know, you, everyone that I've worked with also makes products and gets them out the door and doesn't need us to do that, right? The, the question is, can we dramatically improve the processes, those

Chris: 29:58

I, I joke all the time. I go, the worst thing I can ever do for my own mental sanity is go out on the manufacturing floor and help run machines. And it's not that I'm not capable of it. I do it regularly. I, I, I And I do it on purpose, but it, it hurts me mentally because I see all the inefficiencies, I see all the things that are broken and I see all the opportunities for improvement. And I get overwhelmed and I'm like, Oh gosh, where do I even begin? You know? And so I, I get it. You know, it's, it's hard sometimes when you don't have the right tools that are connecting all the dots to be able to see, Oh, we're going to have a problem here. I would much rather see it five days ahead of time than the moment I need to build the product.

Michael: 30:40

And there's something else, there's, there's something else you mentioned. That's a really good point that I often see as well as even companies who have good digital data, they might have built something with Power BI or something in a command center where they can see particular information. When something is off track, they'll grab a note and they write something down or they send a text.

Chris: 30:59

Yeah.

Michael: 31:00

And so the other part that is critical in all of this is not just the digitization of the process, but the buy in of the people part. Because when we digitize these processes, we generally will bring in the people on the line or all the people who are part of the process to define how they would like it to work. Because they don't like the problems they deal with every day either. They wish it was better. And so you create this change management process by getting them involved in designing a process that's familiar to them because we can digitize what they're currently doing, the paper they're using or whatever, Excel sheets. And, and build in a workflow that then brings the people part along. It's critical. And lastly, is when an issue gets you know, brought up, our system can trigger off a workflow that will send an action item to a particular person or somebody can assign somebody. They get an email, they get notified in their system, they can track everything they're supposed to be doing to address all these issues. Because often this stuff just gets lost.

Chris: 31:59

And

Michael: 31:59

So, and then we'll talk about AI and how that affects it.

Chris: 32:02

how that affects it. So when you build and you talk, making sure

Michael: 32:09

yeah. So. One is the digitization, because it has to be digital, because you can't use AI if it's not digital, but two is building these processes with the team for the change management, making sure that the action item and follow up process works, and that management is bought into the following up. Because people will stop, if you for instance have a 5S process, I don't know if you guys run a 5S process where you're, right? You have a 5S process, if people stop doing audits, people stop doing 5S. Right? So management has to do the auditing, or nobody, nobody continues to do the process. So how does this all relate to AI? So, when we were defining the platform at SAP, it's terrific. They do great branding as well on it. It's the Leonardo platform is what it became after I left. I don't know if it's still called that, but basically we had what they called fog connectors that would connect out at the edge and bring in the MQTD data. They could even do some analytics at that level. And then we brought everything into the central HANA infrastructure that could do vibration analysis, and we were doing predictive maintenance and the traditional AI type of things. And those have somewhat become commodities. You can connect to those on Google or Microsoft or SAP or Amazon. They have their own infrastructure for dealing with IOT kind of analytics around those types of things. The, the generative AI, the new stuff is a whole different thing in that it's large language model based and.

Chris: 33:35

idea behind

Michael: 33:36

The idea behind it,

Chris: 33:38

as aware

Michael: 33:39

you know, and I'm not an AI expert, although I did do a bunch of research on how gradient descent models work and how they figure out the stuff. It's,

Chris: 33:46

the time

Michael: 33:46

way smarter people than me. But anyway, the point for us is, and I run into this all the time when we're talking to people, people want to do it, but they don't know how to pragmatically build it into the system. Like how do we do it in a way that actually drives value for people? So what we've done, is we've built AI into the infrastructure of our platform, into the workflow. And so what that means is you can apply a smart AI language model to any problem you're looking at. So you asked for some very specific. So let's say you're running an A3 process. So traditional A3, I'll bring up one here. I've got on my you know, you have a standard before and after, what are the gains you expect? How are you going to track them? What are the actions? What's the control plan, et cetera. In the AI version, what you do, is you enter the type of company you are, you enter the problem you're having. Let's say I'm having a problem with spoilage. I'm a food company. I'm a spoilage because I'm having intermittent outages in my in my refrigeration system. And I put in all the data and I click the AI help button and it says it could be any one of these six things. What, which one of those do you think it is? Now, I might have been able to figure that out, but research I've seen says AI is two to three times more creative than the average person in an answer because it has the entire internet as a database to look at,

Chris: 35:00

to look at.

Michael: 35:01

So it comes back with, you know, five or six, could it be any of these? And the person can go, I think they can select which ones they think is most likely, they can add text to it, it goes, you know, I did see X, Y, and Z as part of this process. They hit analyze, it comes back and says, here's possible solutions to those problems.

Chris: 35:17

And then you can

Michael: 35:19

And then you could say, well, I'm going to take number three and try that one. And it generates a task list that you can then go assign to your team.

Chris: 35:24

can then go assign to your team. You can then,

Michael: 35:27

You can then, another one is let's say you're getting,

Chris: 35:30

line, you're

Michael: 35:31

You're on the line and you're getting too much too many errors, too much waste. The waste code in our system would bring up a dashboard that would say, hey, you know, you're over your, your your scrap allocation, sends an email off to your management. And then what it does is it, Based on the error code, it comes back with a list of here. Try these five things. What happens today is they click maintenance. I need a maintenance. And then they stop working cause it's, they got to stop. So you have a stoppage and now you're waiting for a person and your maintenance person's over allocated. What we can do with AI is we can bring up here, six things that one, the internet says you should try, but even better, uh, index all of the. Or the documentation on your products, et cetera, and say, here's some of the things that have worked before. Try these, validate that you've tried these, these things before we send a message to maintenance.

Chris: 36:24

and seen things before you send a message to me. SMT pick and place machines, because the SMT pick and place machine manufacturers don't publish anything. They don't, you know, it's all, it's all in PowerPoint and it's all in their user manuals and documentation. It ain't not on the public internet, nothing that, that anybody can scrape and learn from. So it's like, they don't know, you know, we have to, we almost have to teach the AI model, what is specific to our machine and almost have to feed it this information from our manufacturer.

Michael: 37:00

So what we've implemented in the past is systems that will bring in three pieces of data. One is based on the air code, any, literal SOPs to deal with that particular things that are designed. It'll just bring them up, right? And, and so you can take whatever documentation you have with digitize that the other one is to go out on the internet and say, have you tried X, Y, and Z? And then the third one is to index your own databases of what has worked in the past and bring that up. So those are really practical ways that maximize your people. You know, shorten downtime, eliminate waste faster. And you know, there's plenty of ROI models on how to do this. We tend, when we do an analysis to try to have a three ish month return on investment.

Chris: 37:44

Well that'd be nice. That'd be very nice. Now, are you able to find find that kind of investment with a small 50 person business? Or is this, usually you have to be a certain scale to find that kind of return?

Michael: 37:55

no, because, you know, 50 person businesses, you know, depending on the scale of what we're building, we can build things that are highly valuable really quickly and really inexpensively. So like our smallest, we have customers that, you know. Start at 10, 000 and go up to six figures. I mean, we, we, we run the gamut.

Chris: 38:16

Yeah, interesting. Very interesting. Well, maybe we should talk more about business than we should about the podcast. I don't know. This it seems like it might be a fit for Worthington.

Michael: 38:27

let's do it. I mean, you guys will be happy.

Chris: 38:29

I think, I wonder if some of our listeners think sometimes that we have sponsored content. And it's really just that I like to talk to interesting people. And Michael, you know, reached out and I was like, okay, this would be an interesting conversation. I'm not, we're not being paid for this. We just want to talk to interesting people. So it just happens to be a good fit at the moment that we have a particular

Michael: 38:46

Well, you know, the truth is that most manufacturing companies, when I show them a picture that shows a bunch of Excel sheets and whiteboards, They go, yeah, that's, we know that, like that. So it's not that, you know, this is particularly relevant to you, it's kind of relevant to everybody in manufacturing, you know. It's a problem space that isn't addressed and it's not addressed because making enterprise software is hard and if it has to be completely tailored to your business, how do you do it in a 50 person business? You can't, so you end up in Excel.

Chris: 39:15

cater to your business? How do you do it in a business where you can't? So you have to do it yourself. The, the, the big fear that businesses like ours have that, you know, we'll talk about FUD FUD, Fear, Uncertainty, Doubt. afraid that Praxie is going to go out of business and we're going to be left with this, you know, we're. We've built our whole thing on this, and now what are we going to do? We're afraid that, you know, Sarah is going to get hit by a bus. We're afraid, you know, and, and she's our in house support person on, on who's writing all this stuff for us. Versus, everybody knows Excel. You know, and I think that's a lot of times when it comes back to why a lot of businesses like us are hesitant to move to a dedicated platform is, is because of that fear, because of that uncertainty, you know, will it be around is, is Michael Lynch and his team going to be around to support us for the next 10 years when we need them to be, you know, SAP might be very, very expensive, but, We know they ain't going anywhere. You got the U. S. government as your customer. You're going to be around a long time.

Michael: 40:32

Yeah, I think that's fair. One of the things in our system that we've done is, is dedicated to kind of the open ecosystem. So, you have the ability to export the entire data structures out of our system. You have the ability to create a report query, let's say, and export everything into Excel sheets.

Chris: 40:52

Mm hmm.

Michael: 40:53

You have the ability to print out everything in a PDF. So

Chris: 40:56

So if we have to go back to the Stone Age, we can do

Michael: 40:58

You can do it. And the thing is that, you know, when you're a smaller company like you guys are, you're sort of stuck in the place, as you said, and you're even kind of stuck there with your older MES system. Is that what it does for you? MRP, yeah. So, that problem can happen any time, but what we've been, the reason that we thought there was a huge opportunity to be a Mid size and smaller manufacturers is exactly what you laid out is that you're not going to be able to buy SAP. The IT structure is not really affordable. You can try to do it, but you'll be sorry. That's been my experience. And so this is, there's some risk in going with any company, but the risk of not doing it is you just have to stay where you are. And as you said, if you're trying to grow, you've got to find an answer that, that, you know, has the minimal risk that you can have. So we've addressed that by just making our data structures, you know, accessible. So that people feel comfortable that if they ever do need to export everything they can.

Chris: 41:59

you can make good content for the community.

Michael: 42:03

Let's do it. I'll send you a meeting request on Tuesday and

Chris: 42:11

This is awesome. I, I, I feel like I've really you know, my eyes have been a bit open to some opportunities here. I, I, this is something that has been nagging us for over five years. I mean, maybe longer, you know, whenever, whenever computers got over four gigabytes of RAM, that's how long this has been nagging us. Cause that's when we moved to 64.

Michael: 42:32

60, damn 64 bit!

Chris: 42:34

It's literally how long it's been a problem for us. So we, we gotta do something about this.

Michael: 42:39

You'll be surprised. We can help you a lot at a very low cost. You'll be happy.

Chris: 42:43

Cool. Very interesting. Very interesting. Hopefully I can get the PickPlace podcast discount. Is that a,

Michael: 42:48

Absolutely. Absolutely.

Chris: 42:49

Um Well, that's awesome. Is there anything else you'd like to share before we get before we, we get into our pet peeve of the week,

Michael: 42:56

No, I think we covered it all. Yeah.

Chris: 42:57

right, cool. Brilliant. All right. So Michael, if you have one, by all means, we'd love for you to share a pet peeve of yours. If you don't have one, trust us, we have, we have lists, literal lists, plural.

Michael: 43:12

My you know, it's funny because we were talking and I was going to talk about one with sort of business. And then we were talking before about vacations and going abroad and things like that. And this is going to be sweet and sad at the same time. It's people who don't know how to train their dogs.

Chris: 43:31

Okay. Tell me more. Cause I have, I have a puppy.

Michael: 43:34

it's just like dogs are total positive reinforcement animals. Now, you watch the experts. If you have a real psychopath dog, you've got to hold them and do things. But I see people all the time yanking and yanking and yanking and yanking on these poor animals, and they just want to do what you ask them to do.

Chris: 43:51

right.

Michael: 43:51

You've just got to love them. That's all.

Chris: 43:53

I know, I know. Well, so I definitely have that problem. I don't know if I, I normally that they sit their butt down and then they just get really excited. And they're like, please pet me, please pet me at my, my two little pugs. They get there. They're they're easy to train though, because they are so food driven.

Michael: 44:09

That's right. They'll do exactly what you ask them to do. You just got to, and then love them, just give them a treat, love them. They'll do what you say. That's, that's my only pet peeve.

Chris: 44:17

So, are you saying, how does this relate to travel though? Because you mentioned something

Michael: 44:21

Because when I was at, when we were, when I was overseas recently, I saw this guy just, it was so bad that people in the market started yelling at him. They were yelling at him in French. So I didn't get involved, but it was just this puppy. And he like, he didn't know how to deal with it. He just kept yanking it. He was like 25. He's a big guy yanking a dog around. I was like, come on, man. It's a puppy.

Chris: 44:40

Yeah, it's a puppy. You just got to pick that thing up and just throw it in their arms.

Michael: 44:43

Just love it! Just love it!

Chris: 44:45

it. Yeah. Yeah, exactly. Well, I'll tell you, we brought our puppy to a local brewery. They had a half marathon. My wife ran the half marathon and I, I couldn't walk five feet. I could not walk five feet with, without people just being obsessed

Michael: 45:02

Yeah, pugs are so cute when they're little.

Chris: 45:04

Oh my gosh.

Michael: 45:05

cute the whole time, but they're really cute when they're

Chris: 45:07

they're, they're not cute when they're like past 10 years old, they get pretty old, ugly past 10 years old.

Michael: 45:12

Hey! Dog ageism!

Chris: 45:15

I still think they're cute, but you know, I'm a little biased. I'm a little biased. Yeah.

Michael: 45:21

Okay, what about you guys?

Chris: 45:22

Well, I, I had a recent experience. I was, I was in, I was in a restroom, public restroom, and it has the automatic sync thing, and I walk up to it, and it came right on, and I'm like, this is great, and I'm washing my hands, I'm like, this is great, and I turn around to grab the paper towels, and it's still running. I'm like, okay, this is weird. It's still running. And then I walk like a couple more. Well, it normally is, isn't it, Melissa? Normally it's the opposite. Normally you're like, my hands are here. My hands are here. Let's get started. So obviously you can tune these things. There's gotta be like a little variable resistor or something in there to tune these sensors and. Whoever ran this public restroom was fed up with not getting water and just crank that thing to a hundred and just said, if you, if there is a physical body in this room, I don't care if they're on the toilet, turn this thing.

Michael: 46:09

You know what's interesting about that pet peeve? It's probably on both ends of that. Like, if it just runs water, that's dumb. And if it won't give you water, that's dumb too.

Chris: 46:17

I think automatic faucets are just dumb, period. It's like valves work. Humans give humans credit. They know how to turn valves on and off. Just let them turn the valve on and off. Do we really have to like, you know, Like nanny this thing with an automatic sensor. I, I like the push. I like the push and it stays on for like 20 seconds. I'm good with that. That's a good one too. Cause that's, that's a good

Michael: 46:41

Yeah, that's a good one. Then you don't have to touch it. Yeah,

Chris: 46:44

Yeah. You can hit it with the back of your hand or

Michael: 46:47

I was at SAP years ago, and they had this coolest one. This was in the executive center where we would talk with customers. They had a, it was like a trough style faucet. And when the water was cold, it would have red LEDs. I mean, hot would be red LEDs and colds would be blue. And so it would change color based on how hot it was.

Melissa: 47:10

That's interesting.

Chris: 47:11

Oh man, Melissa, we got to add this to our shower faucet. Yeah, this is the ongoing joke with the PickPlace podcast. We wanted, we wanted to invent a product that, that you could plug in the temperature of the water you want for your shower. And it just comes out the correct temperature.

Michael: 47:25

The right

Chris: 47:26

out they exist and there's like, yeah, but now we got to add LEDs to it. Now that's our differentiator. We'll add the LEDs to it. And we'll be, we'll be, forget the Pig Place podcast. It'll be the Shower Faucet podcast. We'll get, we'll get Kohler on, Delta. Who else can we get on here?

Michael: 47:40

They'll all, they'll all own you by then. They'll buy it out.

Chris: 47:43

That's right. Yeah, no, I I no, I w I, I am very frustrated by automatic syncs. I stayed at a my parents have the, when I go to visit, they have the you just tap the faucet and it turns on. You tap it again. It turns off the like touch sensor, the whole faucet itself. You just tap the faucet and leave the valve open. I like that. But the best I ever had, I stayed at a Airbnb once and I had a foot pedal. So as you walked up to it, you press the foot pedal and the water turns

Michael: 48:12

I have one of those.

Chris: 48:13

amazing.

Michael: 48:14

Yeah, they're good. It's on the, at the kitchen there, you know, the, the island has a, has a sink in it for doing vegetables. So they put that in because it makes sense. You have all these vegetables. You got to turn the water on.

Chris: 48:26

and you want, and you don't want to just leave it running and wasting thousands

Michael: 48:29

Yeah. I think they're,

Chris: 48:30

want to turn it on when you need

Michael: 48:31

I think they came from hospitals. I think that's where they're

Chris: 48:34

I believe it. I know, like, I remember when I was a kid in like an elementary school, we had the one where had a big giant bar. They'll go across and anybody could step on it. And there was like a trough and all these little spouts would come out and you'd wash your hands. And I went to a brewery recently that had one of those. And there was a young kid in there. He's like, how do, how do I turn this on? And I'm like, I was like, it's official. I'm old. I'm officially, I'm having to teach the kids these days. They don't know about foot

Michael: 49:01

Back, back in the Soviet era.

Chris: 49:05

We all had to wash our hands together, uphill, both ways. Oh, this has been brilliant. Thank you very much, Michael, for coming on the show. If

Michael: 49:15

of fun, you guys,

Chris: 49:16

what's the best way for people to get in touch of you, if they want to

Michael: 49:19

super simple. Michael dot Lynch, LYNC h@proxy.com.

Chris: 49:24

that's, that is michael. lynch. I'm, I'm. I couldn't, couldn't think of a better email address. That's exactly what I did for my own. Chris. denny at gmail. com back in the back in the invite only days of Gmail. Trouble

Michael: 49:37

you got it.

Chris: 49:37

somebody signed up for Chris Denny with no period at gmail. com. And I randomly get this person's. Emails. And it's very frustrating. And I have a bone to pick

Michael: 49:46

That's a p. Another pet peeve for another podcast,

Chris: 49:48

Oh, that's not just a pet peeve. This is like, it's just like identity theft. It's like, it's like I, the guy, he, he, he's British. I know

Michael: 49:57

Oh, that's even worse.

Chris: 49:58

he's British because, because I get all of his, his emails thanking me for donating blood to the local, like blood bank somewhere in London. Like,

Michael: 50:08

like a great

Chris: 50:09

he's probably a great guy, you know?

Michael: 50:11

Do you get like footballer type emails?

Chris: 50:13

I, I get. Well, so he must be a fan, he must also be a fan of American football, which is not too unusual in the UK, there's actually a fair number of American football fans out there, because he belongs to a, like a Yahoo Sports or ESPN Fantasy Football League, and This is a children's show, but I'll, you know, you're not children's show, but this is, this is a family friendly show. We, we don't, we don't use, we don't use foul language on the show. And this isn't, this is about as saucy as we're going to get, but the name of his football team is suck my footballs, which is, and I, yeah, I got the whole thing. He's and all his, all his. Every other team in his league, they all got these like little innuendo names. Yeah. Yeah. So he donates blood, he plays fantasy football, and I promise I'm not sleepwalking and cause how can I, how can I cross the Atlantic and donate blood? You know what I mean? I guess this is, this

Michael: 51:08

You'd be a really great guy then.

Chris: 51:12

In my sleep, crossing the Atlantic. Nope. No, sir. But yeah hopefully, hopefully you don't sign anybody up Michael Lynch without the period at proxy. com

Michael: 51:21

No, there is nobody

Chris: 51:22

be very awkward. It'd be very awkward. Well, great. Thank you very much for coming on the show. As always, if, if anybody wants to get in contact with us, we're contact at pickplacepodcast. com. You know, there's, You know how to get ahold of us. Just get ahold of us. And, and I got some emails last week. I will follow up with them. I promise if you haven't heard from me, you'll hear from me soon. Thank you very much for reaching out to us. Turns out, Melissa, people love conveyors and they want to know more about conveyors. See, I told you it was

Melissa: 51:49

conveyor's episode part

Chris: 51:51

you it was going to be a good episode. So, as always, please tell a friend if you've enjoyed the show and hopefully they'll enjoy it as well.

Melissa: 51:59

Yeah, thanks for listening to the PickPlace podcast. If you like what you heard, consider following us in your favorite podcast app, and please leave us a review on Apple Podcasts or wherever you get your podcasts from. Thanks so much, Michael.

Chris: 52:11

Thanks, Michael. Thanks, Melissa.