SUMMARY KEYWORDS
Data trust, AI initiatives, data quality, data enrichment, data lineage, data observability, data remediation, data automation, data integration, data trust score, data management, data innovation, data-driven decisions, Industrial Talk podcast, manufacturing.
00:00
Scott, welcome to the Industrial Talk podcast with Scott Mackenzie. Scott is a passionate industry professional dedicated to transferring cutting edge industry focused innovations and trends while highlighting the men and women who keep the world moving. So put on your hard hat, grab your work boots and let's go
00:21
Well, hello, industry professionals all around the world. Welcome to Industrial Talk. This platform is dedicated, as I say all the time, to you, because we need you to tell your story each and every day, because you need to be inspiring the next generation of industrial leaders, given what's happening in industry each and every day, it's an exciting time, and you have to be advocates and tell your story all the time. That's what Industrial Talk is here for, and it's here for you. So you're bold. I said you're bold, brave, and you dare greatly, you innovate, and you are just changing lives and changing the world each and every day. Jay Limburn Ataccama is the company. He's in the hot seat, and you know, we're talking about data. Let's get cracking. Yeah. You need Yeah. How about that? Yeah. Data is always you need data. You need to collect more data. You need to get more insights into that data. You need to you need smart people like Jay and companies like Ataccama to be able to sort of help you along on that journey, just because it's important, big, big time. Now, with that said, there are a couple of things that I want to put on your calendar and make you aware of there we have out on Industrial Talk the Podcast Network. We have two, two new podcasts. Now, one of the podcasts is a really fun one, and that is a podcast called Ask Molly. And the reason I say ask Molly, Molly is my daughter, and you know what I do when I'm challenged with clients that need marketing help, and I need marketing help. You know what I do? I ask Molly. How about that? That's pretty doggone simple. You can ask Molly. So we have a podcast out there. We're talking about one. Do email campaigns still work. There's a debate going on out there, just FYI. Do they really work? You have an inbox. I have an inbox. You see stuff coming up by so how can we take that technology and still make it relevant? Yep, ask Molly. You ask Molly. It's important. The second podcast that we have going is called the Business Beatitudes. I have blogs out there. I have I'm in the middle of writing a book, which is absolutely difficult. God bless the people who write books. My goodness, that's tough. The bottom line is that here in industry, so I always look at industry from the people's perspective, individuals that are in industry, the industry that helps other people solve challenges and become successful in their own right. That's a reality. So I'm always thinking about that. So the business beatitudes takes into consideration these seven components. Success comes through those who are sacrificial, humble failures. You know, you got to fail a couple of times here and there, absolutely calm selfless, forgiving and generous. And then, of course, we're going to expand upon that. But I believe that companies, especially today, really need to begin to embrace those virtues that are really other focus. Already other focus. Let's be more other focus, if I can say it that way. So anyway, two ask Molly business Beatitudes, couple of events that are on the calendar for Industrial Talk, starting in January, we're going to be broadcasting from PowerGen. And if there's a topic, if there is something that is just absolutely on fire, into use on fire when it comes to PowerGen, but you get the picture, you understand what I mean. It's power what do we do? How do we get more out of it? Because of these data centers and and the increased demand for power, PowerGen, January, 20 through the. 22nd this is in San Antonio, and it's being held at Henry B Gonzales Convention Center. The links will be out on Industrial Talk. Check it out. I'll be there. You want to buzz on by and you want to be able to chirp with me? Yeah, I'll be there. The other event that is absolutely incredible is M D and M West. This is an Anaheim, and it's at the convention center. It's February, 3 through the fifth. And another topic, another industry, manufacturing is everything. Just FYI. There's you got you you want to near shore or reshore. We got to talk about expansions. We got to talk about people. How do you fulfill that is automation. There's so many little tidbits of information that are necessary for us to succeed in manufacturing, and that's just global there. How about that? Man? There's a lot going on. I'm all happy. Let's get on with conversation, talking data, talking with Jay Limburn Ataccama is the company. Check them out. Everything's out on Industrial Talk. So you can't, you can't come to me and say, Scott, I can't contact Jay. Well, you didn't go to Industrial Talk, and he's pretty active out there on on LinkedIn, so there's a stat card. So anyway, here is Jay Docken data. Hey Jay, welcome to Industrial Talk. Thank you very much for joining. I appreciate your time. How are you doing
06:35
today? Hey, Scott, I'm great. Thank you so much for having me. I'm excited about this.
06:40
I like the guitars in the background. I have one, but I keep it hidden. It's just like that. But it and I don't play worth a lick. I don't I
06:48
don't know I my three guitars make me look cooler than I actually am, and make me look a better player than I
06:54
see. That's exactly why I have mine back there. I just want to look cool. You know? I there's a, there's an app that Gibson puts out, and it's a, it's an educational app specifically for your mobile, and it's pretty good. So it's, it's the gamification of playing the guitar. I'm
07:13
gonna have to check that out now. I'm gonna go and check that out right after this. Yeah, but I
07:18
got, I got banana hands. So it's, it's more, it more is a frustration than anything. But anyway, it's, yeah, Gibson, it's good. There you go. Plug out. Shout out to Gibson. All right, before we get into the conversation about Ataccama, am I still saying it right, or am I butchering the name? You got it now? You got it? Ataccama? Give us a little background, Jay, on who you are, so that the listeners saying, Oh yeah, Jay's got some mad skills. Talk to us.
07:46
Yeah, sure. So my name is Jay lindburn. I'm the Chief Product Officer at Ataccama. Ataccama is a company that focuses on Data Trust, which is that means we give meaning to your data so you can make better decisions from it. To do that in my role as Chief Product Officer, I've had to pull on pretty much my 25 year background in the data and AI space. So I spent all my career working with data, trying to provide understanding to data for large organizations around the world, and really working in this really exciting space as it's evolved over the last few years to start to support some of those awesome sounding AI initiatives that everybody's talking about.
08:27
Well, I got to tell you, man, it is an exciting time and and everybody's talking about data, and it starts with data. Here's one of the challenges I always when I started, when I did implementations over way over here, many years ago, it was always the conversation around, well, how far back do we go with the data? How do we clean that data? What and and everybody's like, yeah, we gotta, we gotta make sure this data is clean and all of that good stuff. And then all of a sudden, week into that project, everybody's like, let's just go forward. Let's not clean it anymore, because it's miserable. Yeah?
09:07
And, you know, it's so funny, because a lot of the time I've been focusing in this data space, people did things like cleaning their data because they needed to do it. They didn't really know why, and they couldn't tie it to any kind of business initiative. It's just like, we really need clean data. Yeah, but the really exciting thing for me now is, over the last few years, because of the AI initiatives, there's now a renaissance in the data space, because people recognize that without getting your data foundation correct, your AI is really going to be a black hole. And so it's cool to be in data at the moment, because everyone cares about it again, and all these initiatives are now tied to the success of these AI programs, which is great, yeah.
09:44
But are you having conversations around these legacy systems like, Hey, this is, this is our ERP. It's been in place for, you know, 27 years, and and it and, and here's another system, and here's another how does how. How does that account? Just take that, that that legacy, and then begin to bring clarity into that data.
10:09
It's, um, this space never gets simpler. Does it? Because, you know, you've got all these years and years of legacy data to your point, there's always a new cloud data system that gets spun up somewhere. And the truth is, you need to be able to connect to all of it, because AI requires the best information from across your entire business. Make sure that you can feed it the best answers. And so, yeah, the way that we operate an out of camera is we allow you to connect to as many different sources as we can possibly, as we possibly can, and give you a view across all that data. We then provide enrichment around the metadata from those sources so you can start to describe what that information is. We then apply data quality to it. We can capture the lineage from it. We can capture the observability around how the data is moving across your entire real estate, and then what you end up with is a map, if you like, of of enriched information that describe all of your data from your ERP and from the edge and from sensors and from all these different data sources, and bring it together into a corpus of information which you can then use to train your models and provide better analytics. At the end of it,
11:20
how do you because it's still a human equation. So you come to me, I'm a manufacturer. I've got all these systems. I've got all this data, pretty much everywhere. I'm slipping in it, I'm stubbing my toe on it. And how do you bring everybody together to shake that head? Yes, that's the data we want. And yes, that's the information that is relevant to our business, and we like it because there's data. Oh, you got data. It's a tsunami of data. But how do you bring it and say, we'll put that over there, we'll use this data.
11:59
It's really hard because, like, if this isn't just, this isn't just a technology problem, that's the thing. It's a, you know, people are important, then the processes around the people are important, and the technology is really kind of the third pillar from it. I think if you look historically, a lot of these kind of large scale data projects took a long time. It could have taken two, three years, historically, to start to bring data sources in and try to understand that data. And you get through the end of it and kind of like, okay, great. Now we know what we've got. Where's the value? What am I going to do with this data? Whereas now, particularly in our out of camera platform, we've kind of been able to use AI inside our platform, so we can now automate a huge amount of that gathering of information, so it's a lot easier and a lot quicker to get to value. And once you've got everything captured and documented and understood, and then you've got something tangible to show the people and how it can enrich the processes of creating AI, it actually is a massive unblock. So you go from these two, three year long data projects now into delivering value into a, you know, less than a few months in some cases, which means it's, it's much more manageable and justifiable.
13:14
How does, how does? How to comment, okay? Now I've got it okay. I I've got all this data, I'm starting to pull it all together. I'm like, yeah, it's good. And this, this is data is good. This data that I don't care about right now and maybe for the future, but right now I don't care about it. I'm I'm trying to drive to something that is tangible right now. What does at a comma, your Team J do to ensure that the results are what they need to be like, not not false positives, or whatever it might be that, yeah, that's that's really what's happening out of and I trust that. I've got a lot of trust in that data. How do you do that? How do you get to that point?
13:58
You mentioned the word trust there, which we see, is really, really really important to this, because for years, in my 20 years in this industry, people have been trying to catalog their data and explain their data, but it was very, very siloed, historically across lots of different types of data. And what we're able to do in our unified platform is we brought together, first of all, Legacy sources and modern data sources and cloud data sources so you can see all of your data together. But then we've taken technologies across what was previously separate domains, such as data quality or data lineage or observability in reference, data master, data management, and brought all of those capabilities together so that we can enrich and provide you the best view of all of that information. And then what we can do on top of that is provide a trust score, and that trust score gets associated to each of those pieces of data inside the platform. So then you've got a company wide metric that you can use to a tell you how much you should trust that data. And. And then secondly, you can use our platform to improve, over time, the trust in that data and increase that score by doing things like data remediation and improving data quality. And so over time, you can start to not just share how much you should trust that data across across the company, but actually then show that the improvements that you're making by using the platform to increase the level of trust.
15:21
Two questions. What is that? Trust score? Is it like one to 100 100 is great? 25 not good. So is that sort of the scale, or is it red, bad, yellow, then green, good. But truth is,
15:39
it's the truth is, it's completely configurable, and that's really important, because a manufacturing organization may think that a trust score of 70% is fine. However, in banking, if you're doing a credit check, you probably want to, you know, a score higher than 70 and so truth is, every industry and every company is different. So we allow you to configure the thresholds of what is good and what is not good, and then allow you to determine, you know, if it's green, it has to be over 90, and if it's less than 90, it can't be used for a credit check, AI training model, for example. So it really allows you to adjust things specific to your own company and industry.
16:15
Does the solution. Let's say here, here's something that is happens all the time in an industry, I have a system. I'm depending on this person, a human being, to enter data into that system and not pencil whip that data into that system. You know, there's like, I was out there in the field. This is real stuff, and I I have lunch, and so I just got done, I got to just sort of throw that in there. Does your system say, hey, hey, you're I'm seeing some trends here that are impacting the data quality. FYI,
16:57
yeah, many dimensions to that, but kind of just to hit on a few of them. So yeah, we can track the the overall quality over time and how it evolves over time, whether that is degrading over time gradually. But we also do things like anomaly detection. So if someone fat fingers on the keyboard and suddenly, you know, they put in, I don't know, an invalid state code, and we've now got 51 states in this data set, for example, then that anomaly would be automatically flagged up for the system. And then you can configure thresholds to say, well, just auto correct it, because we can use our own inbuilt AI to auto correct those issues. Or maybe you want to go and flag it up and put a manual check in place from an individual put it onto their job queue and allow them to manually go and make those changes.
17:40
Yeah, you want to compress that time. You don't want to let it go. You don't want that, that, whatever, that anomaly, to continue to sort of, Oh, yeah. I just didn't just keep on going, because then that defeats the purpose of
17:51
of trust. You got to move things to the left in that supply in that pipeline, because, you know, if you let it go unmatched, it could end up affecting your inventory system or your supply chain, or maintenance on some kind of machine, and the sooner you can catch it in that pipeline of data, the less impact there is downstream from that.
18:11
See, I would imagine and correct me if I'm wrong, and if this is and this is in the world, I hate to hate to say it IIoT, I'm pulling data. I'm I've got a device out on this asset, and this assets vibration, you know, pulling in the data. I would. I want it to the point me that when I look at that dashboard, whatever it might be, I hate to use the dashboard. I don't know that that computer screen, it says, okay, I'm okay. I got to go out there. I know what to do. It does this. There's a lot of automation and a lot of proactive push, of saying, right? Create a work order. This. That the other thing so it makes me, the human side, more efficient, right?
19:00
Yeah, it does. I mean, it really does, because you're able to capture more information, but then you'll start to be able to actually trust that information in the first place. The amount of people, the amount of dashboards that get created across different organizations that are actually based on information that they don't trust, making important decisions upon them is scary as hell. And so the more you can do to bring the information together, enrich it with all the information that you can and then feed that into your AI models, that then feeds back into that predictive maintenance, or whatever. That's the unlock, and it it becomes the cornerstone of automation. If you can't, you can't automate in your processes without being able to trust the data behind it. It's scary if you do,
19:46
yeah, I and it doesn't take long. So if I have a dashboard and I'm pulling in data, and I don't trust that data, I I'm more inefficient. Let's put it that way, I'm not going to trust it, and therefore, I'm going to go out to the warehouse. I'm going to go find and see if. That in for that that part is there, or whatever it might be, but I'm wasting time. Do I go out to that asset I don't trust what's coming in? Yeah, trust, boy. Trust is everything. It's if
20:13
you think about it, if you think about it from a competitive perspective, right? If I'm, if I'm in the, in the industry of manufacturing and my competitors are trying to automate as much as they possibly can, every single process, every single supply chain, every single manufacturing line. The more automation, the lower my costs. And if my competitors are able to do this better than me, because they understand their data better, their data is more understood and enriched and trusted across the business. Their dashboards are more accurate, their AI is making better decisions, then they're going to beat me regardless of how good my products are. And so it's really, really important that you focus on getting that data right. How
20:53
do you address the evolution of data? Let's say it's not static. Let's let's just say, Okay, we we've got a benchmark. This is where we're at. We trust this data. It's right here. But eventually, as time goes on, we're going to continue to pull more data, great, create more insights, all of the stuff that that. Does it ever stop? Do you ever get to a point where, hey, thumbs up, man, we're, we're there. We're, we're humming on, you know, all cylinders.
21:24
I think that's one of the biggest challenges that organizations face, is that you're never one and done right if you think of data as the effectively the blood pumping through through the veins and arteries of your organization, I don't know it's a pretty good idea to keep on top of your blood pressure and make sure that your your blood is pumping well for your health, right? And it's true, it's the health of your of your business. And so you can never take the eye off the ball of how you're thinking about the data and the data's health. And data is always growing. And so you got to make sure that as the data volumes increase, you're applying those data programs and Data Trust programs across that data as it as it starts to grow.
22:01
So with that said, the change management component of this is like, Okay, we got we're here. We're also pulling in data over here. It looks like it's got some quality to it. How do you work within an organization to start beginning to incorporate Yeah, that data is more important than that data, but we like that data. It's a dynamic environment, but it's it requires, again, organizations to say, yeah, yeah. We're good that, yeah. That makes sense. I can't wait for that.
22:36
Yeah. It always changes. Like, what's I think that's one of the reasons why there are our Data Trust index has become so popular amongst our customers because it allows you to constantly evolve how different data is used. You may have some data that perhaps has not been very well understood and therefore isn't used by the business at all, and therefore it's not valuable. But you know what, if you start to actually use some of our AI tools to document that data automatically, and then it starts to drive consumption. And then you can start to look at the observability of that data and realize that actually that data is now feeding into a report that our CEO uses to the board. Well, actually the value in that data must be pretty high if it's being fed up to the board. So maybe I'm going to use that data for my analytics project or my AI project. And therefore, over time, that data becomes critical, because people start to trust it, because it's well documented and it's well understood, and people know where it's come from. And so it's important to make sure that you're taking a consistent look at that data and using technologies that are automating the processes behind data management itself.
23:47
I think correct me if I'm wrong, Jay, but I think the scuttlebutt in industry is always around that silver tsunami that the individuals that are leaving that knowledge that's in the head of these individuals departing the company. But I think if companies are aggressively pursuing clean, clear data, then that transition, that pain is minimized dramatically, because and as well, as you know, the the big thinker up in the top floor is also retiring, and then we're going to have a different, you know, person with a strategic view of the you know, how to move forward, but, but the, the foundation of that is clean day and and that transition is minimize the pain, because it's always now, it's all painful, it's all yucky, and nobody likes it. Nobody, I don't know anybody does, yeah.
24:49
I mean, you're hitting on the you're hitting back on the culture piece a little bit, right? And the tribal knowledge that exists amongst amongst people. That tribal knowledge exists because even. It's in there. It's either in the person's head or it sits on a spreadsheet somewhere on that person's laptop and changing the data culture across a company, whereby the best information is available across the company. It's well understood. It's got a trust score assigned. It's available through self service across entire lines of business, through, you know, highly governed, highly curated data products which can easily be used. That's that effectively. That's what chief data officers trying to are trying to enable these days across these companies. And to my earlier point, that is what then feeds into kind of the data innovation around analytics and AI use cases. I
25:39
think it's a non negotiable. I think, I think if you're a company that is not seeking that trusted data in your organization, I think you're going to be left behind, and I don't want you to not succeed. So that's that's important. Speaking of the future, put your future hat on Jay. Where do you see it all going, you know, this is, this is a fast moving world. And, and everybody's there is that push to say, I could get that data, I can get that data. And, and they're in that sort of frenzy mode. What do you see? What? What's the what does that future look like?
26:19
Yeah, I think data is the difference between good AI and bad AI, bad AI. The companies that get their data right will be able to do good AI and be, you know, actually roll it out from prototype into production and in, you know, all in manufacturing organizations, for example, that's what's going to lead to successful automation, automation and lowered costs. So I think the focus on data, I mentioned the renaissance in data, that's only going to, only going to increase making sure that, you know, organizations can automate their business processes, automate their supply chains, automate their manufacturing lines, and then automate but Augment, right? We're not talking about taking humans out of the workforce. I don't actually believe in that. I think there's always going to be, you know, human in the loop, operators, engineers, co pilots that sit alongside, sorry, people that sit alongside these co pilots to be able to provide the checks and balances as needed. But AI is going to be able to automate a huge amount of the mundane work and make that make our lives much easier, so that we can unleash the potential we have as humans to focus on.
27:28
Now, you're spot on. I agree with you 100% the reality is, is that resilient. Businesses are constantly looking to be more efficient, and you cannot have an efficient conversation without the ability to be able to extract the data, make decisions quickly, automate, where you can use the human side appropriately. Whatever it might be. I agree with you 100% there should be no fear. It just is is good, and businesses that do that well, you're doing it well. Businesses that are reluctant to do that, it's going to be challenging. I just that's that's how I feel.
28:11
Yeah, I agree. And it's funny, because you see, you've seen all this spending and focus towards AI initiatives lasting over the last few years, but increasingly now people are realizing that that spend needs to shift back to data. Go, get your foundation right. Go build the trust in your data, and the AI stuff becomes much more straightforward. Yeah,
28:30
you're spot on, Jay. How do people get a hold of you? This was a great conversation. I know that we're just scratching the surface, but how do people are saying I need to know more about at a camera?
28:43
Yeah, sure, so Ataccama.com is obviously the straightforward way to get to our company, but I'm available on LinkedIn. Shoot me up on LinkedIn. Outstanding.
28:55
This was a great conversation. I can geek out all day long on it, but I try not to. I try to let people like you geek out better than I do. You are wonderful. Thank you, Scott. All right, listeners, we're gonna have all the contact information for Jay out on Industrial Talk. So fear not reach out to him. Find out more, get your data right. That's what you want. That's what creates a resilient business. That's what is success means. So anyway, we're gonna wrap it up on the side. Stay tuned. We will be right back.
29:23
You're listening to the Industrial Talk Podcast Network.
29:33
Yeah, if, if you don't think that data's not important after that conversation, well, yeah, it's important. And you have this technology stack that's taking all that data, you need to bring it in. You need to be able to create some context around it. That's what had a camera team. Jay Limburn, you know, right there, I got the stack card out there, so that's what, that's what I'm pointing out. Is if you're looking out on the video anyway, great conversation, absolutely wonderful. Appreciate Jay, definitely sharing his insights. All right, we're building a platform that is specifically for you, telling your story. We need to inspire the next generations. We need to be able to do that effectively, and because manufacturing, because data, because everything's a power, you name it is cool. Tell it. Let me know. Go out to Industrial Talk.com. All right, be bold, be brave. Dare greatly. Hang out with Jay, and you will change the world. We're gonna have another great conversation shortly. So stay tuned.
