Discussion Topics

• Knowledge management has always been challenging. What are some of the technical, operational, and cultural barriers?
• Focusing on construction and manufacturing, in what areas are we most concerned about tribal knowledge and losing top people?
• What are the opportunities to close the knowledge gaps with technology and to better support AI initiatives?

Participants

• Isaac Sacolick
• Bob Salaj
• Joanne Friedman
• Heather May
• Joseph Puglisi
• Martin Davis
• Derrick Butts
• John Patrick Luethe
• Liz Martinez

Transcript

[00:00:01] Isaac Sacolick: Greetings everyone.

Welcome to this week’s coffee with digital trailblazers. This is episode 143 for us. We are going to take our usual slow ramp up to get everybody here on board and excited to talk about our topic today, the real cost of tribal knowledge. Why losing your people can bring OPS to a standstill.

This is a conversation that we’ve actually danced around a few times at different coffee hours talking about how important it is to work with our staff and capture knowledge, create processes, move away from gray work. And I’m really excited to have Bob Salai here with us today as our expert and we’re just going to give it a few more minutes to everybody joins. I actually clicked the Go live button a little bit earlier than usual just to see what would happen. It’s always fun to Mess around with LinkedIn and make sure everything’s working before you get started.

But really happy to see here folks. Say hello in the comments stream. Steve, thank you for getting that started and do want to hear your thoughts and your questions around knowledge management, around knowledge sharing and how we create processes. Hi Kristen, it’s good to see you again.

Hi David. Kristen and I saw each other just a couple weeks ago. Kristen, I’m going to be right near you Oddly I’m going to text you later on my drive back from Tucson, Arizona back to San Diego. So Valaba, good to see you and everybody just say hello in the comments stream.

This is an event I love to see people participating in especially this week. We’re talking about the real cost of tribal knowledge while losing your best people can bring OPS to a standstill and for everybody’s knowledge. Today’s episode is sponsored by Quickbase. Quickbase is an AI powered operations platform that is designed specifically for your unique processes and it’s not a one size fits all approach for those of you who are long term watchers and participants. You know we had a really interesting conversation, I think it was two or three months ago about the problems with Saspral and all these point solutions, the hundreds of different applications that exist in many enterprises and the impact of those and now today we’re almost talking about the opposite of this.

What happens when we don’t have processes, what happens when we don’t have documentation and over time it becomes a debt to the organization. We cannot easily onboard people, we can’t easily improve processes because you talk to one person and they know their little bit of what’s going on and so forth. So today we’re talking about the real cost of Tribal knowledge why losing your people can bring OPS to a standstill. I want to start with just an overall discussion around the knowledge management challenge and get some thoughts from everybody around what are some of the technical, operational and cultural barriers that prevent knowledge management from happening? I’m going to start with Bob. Bob is our expert from Quickbase. Bob, it’s good to have you here and just tell us a little bit about what you do at Quickbase and then share your thoughts around some of the challenges around knowledge management from a technical operation and cultural perspective. Welcome, Bob.

[00:03:45] Bob Salaj: Yeah, thank you, Isaac. And thank you to everyone here for joining us here today. It’s really exciting. Thank you for having me back. By the way. It’s always a pleasure here to join and to talk through this. And tribal knowledge, it’s always been really the hot topic, I think, of my career just to kind of go back, really. I’m a principal industry advisor here at Quickbase, primarily for the construction industry and really my career has spanned around lean continuous improvement. Starting off in manufacturing, going to Wesco distribution for electrical, distribute, distributing, but really for the last decade in construction, have been leading up innovation teams to kind of really look at their processes and how are they rolling out digital solutions for not just the office but also the field kind of going into this. So quite often focus a lot on lean continuous improvement, but use tools like Quickbase to help fill in those gaps and really kind of bring a lot of innovation to life in real time. And that’s really, I think I’ve seen a big change for it. So I see a bunch of hands being raised here. I want to give everyone a chance, but just to kind of like kick us off and talk through maybe one of the instances. Because culture, tech, operational, there’s a lot in each one of those and culture I think is the one that I really, really wanted to kind of just speak to first.

Really. I think before we even talk about like tribal knowledge and stuff like that. I’m kind of curious on the organization if they’re primed to receive feedback.

Basically, what is the culture at its heart and are they open to receiving feedback or are the employees feeling reluctant to kind of give their ideas for fear that either it might be misconstrued as complaining or maybe they’ve actually done it quite a lot of times and it just goes on deaf ears and no changes begin to occur.

So what happens a lot of times is that power is knowledge type mentality and they begin to hold on. So that’s something that I’ve seen across the board with a lot of organizations and curious from the rest of the group and panel other areas there, if they’re seeing that as well.

[00:05:52] Isaac Sacolick: That’s a really interesting point, Bob. I never thought of it in that order that you have people who are willing to volunteer information, who want to share feedback, want to improve things, and then over time, when they see management isn’t responding to that feedback, they just stop talking and they just do their own things. I never actually thought of it that way.

Let’s bring Martin in. Martin, have you seen it?

Have you seen that progression or where does it start from? What’s the underlying problem that prevents organizations from really creating a culture and processes and technical capabilities to share knowledge? Welcome back, Martin.

[00:06:33] Martin Davis: Hi, good to be here again.

I think there’s a couple of different things. You know, I agree culture is a big part of this.

You know, if I think back to some of my days in large organizations and part of my background is Ford Motor Company there for many years.

And there’s a couple of things I remember from some of those aspects of where people.

Some of it’s change fatigue as well because they see so much change, so they just kind of get fed up of it and don’t want to share stuff. So there’s too many initiatives kind of drives that. And it’s that cultural thing as well. But there’s also that fear thing. Yeah, the whole knowledge is power.

The person that’s been doing it for so many years has all of this knowledge about how it’s done and then what they’re scared about actually sharing some of that knowledge in case they’re no longer needed.

And I think that’s a mentality definitely for the. Maybe the older generations that are out there. Maybe not so much true with some of the later generations coming through. In my experience, others may disagree with that, but I think it’s kind of that whole thing around trying to help people to understand that for the good of the overall organization and it will actually help their job be easier as well. So kind of It’s Change Management 101 forms a big part of that whole knowledge sharing.

[00:08:02] Isaac Sacolick: You know, Martin, I think you bring up a good point about change fatigue and overload of information. I think that it’s almost like there’s too many voices in the room.

Right. And it creates this atmosphere of when do I get to speak up and when do I get to share?

But I think this, you know, I’m kind of interested, Derek, what you see here, you know, I told a story in Digital Trailblazer about walking into a broken ETL process. And when I went trying to discover what the root cause is, I found a stack of code that took an entire brick of paper to print out. And I did that to go back to my management team and said, look, it’s going to take us a long time to unravel years and years of, of engineering work that now only one person in the organization needs, understands how it works in this stream of undocumented code. We’ve all seen this before. So Derek, you know, what have you seen as you know, some of the technical and operation and cultural barriers that prevent knowledge sharing from happening?

[00:09:08] Derek Butts Yeah, you’re spot on with the, the code based stuff. I’ve actually seen it in industries where, you know, people are doing code, they’re developers on site and all of a sudden they leave and only to find out that they were knowledge hoarding. So they were holding the code, they did not document it properly, they didn’t document how different scripts were written and why they were written. So instead of the one person now having to go back and try to figure out what’s going on, it took several people to try to go back and figure out what’s going on, which again is going to delay progress in the company. It’s going to delay efficiency and Alfaro is going to, it’s going to extend the amount of money that company can generate revenue wise because now they have to go back and fix a product that they thought they had access to when they realized they don’t. I feel the lack of awareness with security associated with that is the biggest point because there weren’t enough security measures put in place to not allow that from a cultural barrier point of view. I look at other things when you talk about operational barriers, you know, these become single points of failure when you have a key person holding on to certain information that’s undocumented.

It affects different things down the road. Vendor relationships, it affects contract relationships. All these things are ripple effects across the organization which cripple the organization in that particular sector. Or maybe overall these are kind of things that we need to figure out. And the other thing I’ve seen, there’s not enough cross training.

One of the things I focus on when I go into organization is say I need to make sure the top level person knows what the bottom level person and the bottom level knows what the person below them. So if one person should be out for any reason, the show doesn’t stop, the show continues to go on. And this is important for the continuity aspect. To make sure there are no barriers to entry would need to do things like that. But the other thing I look at too is just the, you know, poorly integrated security tools are not actually addressing these things across the board.

If they’re doing it properly, they should be catching these things and labeling documentation, see how it. And you know those people that are looking at sending information to their personal devices as opposed to using business devices, that was not a big thing I saw in a lot of different areas. And it was okay because they didn’t have the proper systems in place to track it. And when you look at that across the board it just creates multiple barriers for management to kind of keep up because they don’t know one, it exists and two, they don’t know what they need to do to fix it immediately.

[00:11:23] Isaac Sacolick: Spot on, Derek. There’s also a comment here from Vape Hub in the comment stream. He says I think it needs to be top down, top down driven.

Liz, why are we so bad about this governance? I mean whether it’s security or operations or code, an end to end process. I mean why aren’t leaders demanding that we have good documentation before we deploy?

[00:11:50] Liz Martinez: Well, I mean why aren’t leaders demanding more documentation? I mean, I’m going to say something really simple and that’s because they’re cheap. They want, they want it, they want it now, they want it yesterday, they want it fast, they don’t want to spend any more money. And like documentation is seen as some, you know, basically a waste of money. It’s, you know, it’s a barrier to, you know, revenue in some ways overhead.

What they don’t see is that it’s actually if they don’t do it, what it’s going to cost them in terms of risk and maintenance long term.

I mean I’m going to talk about something that’s really basic which is, you know, an ounce of prevention is organizing, organizing your documentation. A big people have spoken multiple times on this call already about knowledge hoarders and you know, people in fear of losing their jobs and you know, all kinds of things like that. But really just having basically organizational of your document, of your documentation so that people know where to find things. I love what, I love what Derek said about cross training and making sure that you have redundancy or at least you know, some kind of succession planning so people know who’s second in command if somebody is God forbid, hit by a bus or out sick or whatever. But knowing where the documentation is is huge. Just knowing where it is and nowadays we should be thinking about implementing small language models so that we could drill into our documentation very easily. There’s no reason why we shouldn’t be able to find things super quickly if somebody is out sick.

So I’m a, I’m a big believer in making sure that everything is not only well documented, but really easy to find.

[00:13:45] Isaac Sacolick: Liz, I’m going to disagree with you in one aspect. All right.

[00:13:49] Liz Martinez: Okay. All right.

[00:13:50] Isaac Sacolick: Yeah. And then we’re going to let Joanne mediate this. I don’t think leaders see it as a waste of money. I think what happens with leaders is, you know, documentation and process creation is, is essentially an investment for the future. Right. And a lot of them are forced to think short term benefit, which you brought up, and are forced to think through what all the stakeholders are asking them to do and trying to get as much in there and so they don’t make that investment in the future. I think part of the issue, and I’m kind of interested in Joanne’s perspective on this part of the issue is we ask for too much complexity in our implementations.

Right. We need. And we have too much over engineering of our requirements. We have too much over engineering of our applications. I think as we start heading into agents and small language models, I think there’s a risk of that happening.

Another round of this where we wanted to do 100 things instead of doing 10 things well, and that drives either a complexity of process data or implementation and sort of looking for, let me just say lower Cody, ways of doing things where the natural expression of what you’re building is self documenting. Joanne, I’m kind of interested in your opinion. This, you know, it’s not worth it. Or are we just doing too many things wrong up front that make and create this operational and technical and cultural barrier around knowledge management?

[00:15:27] Joanne Friedman: I think we’re doing it wrong.

[00:15:30] Isaac Sacolick: There’s that mic drop I told everyone.

[00:15:35] Joanne Friedman: I mean, I wrote a very, call it a book several years ago now on knowledge management for the Conference Board of Canada. And one of the things that I sort of put out there was we have to look at knowledge transfer and knowledge management as in a different way than we currently do. Which, you know, from the IT perspective it’s about documentation, but it’s really not. It’s about community.

It’s about getting people over the notion of if you don’t keep your knowledge to yourself, you’re going to lose your job. And rather make it about the idea that knowledge sharing is the same as learning. And if we’re continually learning, we’re continually growing and that that brings benefit to everyone. So if you look at it from the what’s in it for me kind of thing and change the culture and incentivize people to share their knowledge, you’re going to have a different mindset that approaches it from the get go. So that could be leadership saying, hey, you know, we’re aware that like in manufacturing now, 33% of the workforce is going to retire in the next two years.

Oh, gee, we have to pay attention to this. How do we make sure that the tribal knowledge, or rather more appropriately termed institutional knowledge, is valued as an asset and does not walk out the door with the people who are retiring? Because then we’re really going to be SOL when we need a fix from somebody. Think about a maintenance engineer or someone in a factory who can walk down a line and knows just by the sound of the machine that something is wrong with it because they’ve been doing it for 25 or 30 or 40 years, that knowledge needs to be transferred to someone else. And the way to do that really depends on how you incentivize the individuals, how your culture allows for that incentivization, whether it’s, you know, a gamification or a reward system or a, hey, you get a gift card from Amazon, whatever it is, how that starts to flow should be much more conversational. And to your point about agents and agentic AI, absolutely. You can look at the prompts that people use. You can store the data that’s in the prompt and the data that is returned by the system as part of that knowledge management platform and then use it to inform learning systems, inform training systems, and also take it even one step farther and inform leadership about.

Did you know that these people are really, really subject matter experts in a way that you don’t expect them to be? In other words, let’s get away from the code and away from the paper and the documentation and have a system that allows people to contribute with an incentive without fear.

Agents can do that.

[00:18:48] Liz Martinez: I would have. Since you’re going contrary to my conversation about being cheap, I would have to.

I’m going to come back and say, first of all, Isaac, I think that what you said about more complexity versus simple is actually exactly the same point as I made, that they would rather invest in complexity than in documentation. So it’s actually the same comment as I made. But when it comes to what Joanne said, which is amazing, I would love to have that major culture change, it’s almost impossible to pull off. So in a situation where you get steering that Titanic away from the iceberg, what can you do?

And I would say organizing your documentation and getting it to the point where you can actually find things and is at least, at least something that’s tactical that you can do in a short term.

[00:19:43] Isaac Sacolick: So I want to bring this conversation into two industries that are really important around us. Right. Because we’re talking, when I think about the industrial applications in construction and manufacturing, technically lagging industries for the most part we’re not talking about digital experiences and workflows, we’re talking about hands on equipment.

There are safety issues with that.

These for the most part are slower changing industries in terms of best practices. There’s a lot of built up institutional knowledge around how to hang drywall, how to pour cement and things like that.

So I want to bring this into this fold of real world problems. It’s not just a knowledge sharing problem. This is how do we continue to build our expertise and how do we build processes and use tools to do more of the work that is in the physical world. Bob, I want to bring you back and then go to Joe, John and Martin after that. Bob, talk to me about the construction manufacturing fields. What are we most concerned about when we think about losing people and the knowledge that we’re trying to retain with them?

[00:20:59] Bob Salaj: Yeah, absolutely. And Joanne, I think you hit on a story that is all too really familiar with the construction industry. Right. But there’s a point here that I just want to kind of take one step back on and it’s just been out there because when we think about the term tribal knowledge, the very first thing that I’ve been asked to in a multitude of organizations is labor shortage. Let’s prepare for retirement. Let’s go out and talk to the foreman, let’s go talk to the office personnel and let’s start documenting processes.

But I also don’t want to forget this is not an age conversation. This is not about just the retirement. This is about your tribal knowledge of really the people and bringing it. Isaac, to his point.

Construction is starting to move forward with a lot of innovation around prefabrication and VDC and a whole bunch of models, twin, digital twins, etc. If you lose those people who have tribal knowledge irregardless of age and you’re not documenting in SOPs or whatever else you might have with that, you’re going to be impacted by it. And so when I think about construction, obviously our minds go to the foreman who can walk around and through smell and touch know exactly the percent complete of that particular job and where it’s going to be. But let’s also not forget about the innovation that’s occurring extremely rapidly and that it’s all about this culture. No matter where you’re at in terms of your tenure at the company overall.

[00:22:31] Isaac Sacolick: Let’S bring in some more expertise around this. Joe. Manufacturing and construction, long time working in this space.

How do we change a culture? Process technology in a place where like, like Bob said, there’s a lot of innovation coming.

[00:22:50] Joe Puglisi: I think you have to address some, some myths. And there’s a certain cultural arrogance in employees who say, I just know how to do this, or it’s too hard to explain, it’s too complicated. You’ll never understand.

You know, the fact is, we had, in one instance, we had an ETL process that took more than 24 hours, and it had been built and managed by one individual over a number of years. And so it was sort of sacrosanct. This individual knew it, no one else knew it, and everyone was afraid to touch it because if you broke it, the company would grind to a halt.

If you bring in some young talent and you explain what this process is supposed to do and you reinvent it and it runs in near real time, you have solved a huge problem.

So no one has knowledge that is completely unassailable or irreplaceable. That’s lesson number one.

Lesson number two is there’s a cultural arrogance at the industry level.

I worked for a construction company and they lived on a report called Anticipated Cost Report.

I then went to the reinsurance industry and at a meeting, there was a conversation about ivnr.

And I had no idea what IVNR stood for and even less knowledge about how you would create this thing.

So I stopped and asked, what are you guys talking about? Explain this to me.

And fundamentally, it was the ACR report. It was exactly the same concept.

So, you know, two. Two notions. No one has knowledge that. That can’t be deciphered or replaced in some way, shape or form.

I’m sure somebody out there is going to invent an agentic system that sort of becomes your apprentice and learns and replicates your processes. Right, Joanne?

[00:24:48] Joanne Friedman: Yes, sir.

[00:24:50] Joe Puglisi: And as far as, you know, industry uniqueness, I see this particularly in the hiring process. You know, if you haven’t worked in pharmaceuticals, you can’t work in pharmaceuticals. I call BS on that.

They all face fundamentally the same issues.

On a technology front, yes, there are some nuances, I admit it, but there’s a lot of commonality.

[00:25:14] Isaac Sacolick: Thank you, Joe. Let’s bring John in and Martin, we’re talking today about losing people and working through tribal knowledge. And I love this cultural arrogance versus learning culture. I love this idea of industry jargon as a killer of culture.

Bring us down to earth, John. Why are we having such a hard time when we get into agile teams and development teams and we’re building applications out, we’re building agents out.

I had some developers who told me the code itself is documentation.

I’d love to hear your perspective on this.

[00:25:54] John Patrick Luethe: Yeah, and I, I agree with almost everything that’s been said today, but the, the place that I think really has a lot of the blame is the management. And I just, I haven’t heard enough of the criticism of the management, which is, first off, if, if you’re going to have people tell people do documentation, like unless if you’re not going to go out and audit it, you really don’t know if the, the documentation is good.

And so what I’ve seen in companies that are really healthy, My last company, we did light manufacturing, light assembly, we, we would keep track of what people had, what skills. We would test all the skills on a yearly basis. We would keep track of like what business processes people can do and we would have people do different, different business processes over time that would allow us to have people a lot more fungible and we would systematically keep track of like the, the skills that people had in, in a database. And we would have people rate themselves and we’d have the managers review it on a yearly basis and we, we would look to see like when somebody left, what kind of skills are walking out the door.

And so I, I just, I agree with everything that people said. But, but like if you want to have good knowledge management, like there’s a ton of responsibility that goes to the management team. They have to make sure that different people are doing things, that people are cross trained. They have to audit the documentation. They have to have people that are unfamiliar with something run through the documentation. And if you do those types of things, you’re going to be a lot better set up if bad things happen in your organization.

[00:27:25] Isaac Sacolick: Martin, you get the mic, but you’re going to have to answer my question. Everybody’s pointing the finger at management. And you sat in the seat of cio, you felt the pressure, you still sit in that seat, you feel the pressure to deliver and improve.

And how do you handle this? Knowledge transfer, culture, teaching, learning, documentations, process cleanup, all the things that make us more operational, resilient. And I’m throwing this out so that Derek can also comment on it all these things and we’re not doing it as leaders.

[00:28:02] Martin Davis: Well, I think there’s only two examples and then a kind of a CIO perspective as well.

So two examples, both come from the manufacturing world and one is kind of a contrary to Joanne’s point earlier about incentivizing people to share knowledge. So I was involved in a company and there was a team of four people, three senior analysts, one manager.

And they had been looking after this specific system, a very important system, for 15 plus years as a team.

And there was significant issues of groupthink as they thought and acted as one body. The four of them, they were all heading towards retirement age and they did things their way.

And everybody above more senior management thought everything was fine because the system ran every day, there was no issues, etc.

All four of them took an early retirement package and left on 31st of December, one year.

That system didn’t run properly again for three months because instead of actually fixing things, they spent all of their time daily tweaking the data. So it ran. So sorting out small issues with the data rather than the system actually dealing with anomalies rather than fixing root cause and things like this.

So they were not interested in sharing knowledge, however much anyone incentivizes that was their way of doing things. There was no interest how you can incentivize them as much as you like. It wasn’t going to change. It was. This was their world, this is how they did things. So that’s kind of one example of where things, you know, that cultural thing don’t always give you an answer.

[00:29:49] Isaac Sacolick: I love that story, Bart. I just do.

[00:29:53] Martin Davis: So after somebody else came in and they reworked some of the system, made changes, the system ran with half a head, looking after it forevermore with very few problems.

So just that kind of where they weren’t letting the knowledge out. So just a kind of a great.

[00:30:12] Isaac Sacolick: Example, Martin, is the lesson learned here, like worst case, you’re going to have three months of pain when the people leave, or is the lesson here like management has to get into the weeds a little bit and see how things are being fixed.

[00:30:31] Martin Davis: I’m not sure how much management could get into the weeds because they close the doors around them. Yeah. And keep in mind their manager was part of that group. It wasn’t just the analysts, it was the analysts and their manager. So the outward facing thing, everyone was rosy.

[00:30:48] Isaac Sacolick: I think that the words you used here, this notion of group think as a barrier, I mean that’s what I would be smelling for, is the Leader saying, have they, you know, has this group created its own, you know, island of knowledge, island of operation? You know, they know how things work. Nothing goes out of this wall. And if you start sensing that maybe you do have to get in the weeds and get involved with this. Heather, I’m gonna. And Derek, I’m gonna bring you back in in just a second. I just want to welcome everybody to our conversation this week around tribal knowledge and why losing your best people can bring OPS to a standstill. Fantastic conversation around manufacturing and construction, in particular about the culture, processes and technology. We’re going to talk a little bit more about people and technology next in our second half and just stay tuned to learn more from this great group of folks talking about where we need to to lead our teams differently around simplifying and creating knowledge for our subject matter experts. Today’s episode is brought to you by Quickbase. Quickbase makes it simple to bring together all your data and teams into one centralized place. With Quickbase, you can collect and connect critical data from various sources into one location, create automated workflows based on your existing business processes, and get real time visibility into what’s happening on the front line. Customers use Quickbase for a variety of use cases like asset management, quality and process improvement, and field service management. I want to thank Quickbase and Bob Salai for joining us today as our speaker from there, talking about knowledge management, process reengineering and our steps to get toward AI. Bob, I want to bring you back in and talk about this because you said, look, you know, the technology is changing rapidly in the industrial space. We’re moving from industry 4.0 to 5.0. We’re bringing AI in and our AI is only as good as our knowledge. So let’s talk about some of the opportunities to close the knowledge gaps with technology and to better support the AI initiatives that are going to bring that culture together instead of separating the culture out.

[00:33:16] Bob Salaj: Yeah, absolutely.

And we’re seeing really the technology kind of move us into another era of really, how do we use AI to look at our entire process, not just the data.

And so we talked a lot on this call, really from the very beginning about standard operating procedures. And I could almost spend an entire four hours just speaking to that because that’s really the foundational thing of auditing the process. And I forget exactly who brought that up. But quite frankly, just to kind of camp on that instance, for one part, everybody goes in and says, let’s go create a standard operating procedure. And it’s almost like checking a Box in the moment we create that document, it’s almost kind of a myth. It goes into the ether and no one actually ever pulls it out again because no one’s auditing it.

So I use Quickbase to actually create an application so that we can actually have the dates and lessons learned of what is actually occurring with that SOP on a regular basis in terms of one instance so that you are actually going out and you’re setting a timeframe of six months, nine months, whatever it is, to keep your operations up to date because management may not have the time to go into every single process. Right. But we need to start using this technology to final the data into a centralized location.

Using Quickbase, you think about that for your estimating department. You’re capturing all the standard operating procedures for that and you’re going through all this time and effort for it just to die on a vine. But all of this information has to come back into one instance to educate our language models, to educate how we actually use AI to audit the process and find the gaps that aren’t documented at the end of the day. So it’s going to help out with ramping up. But to me it’s all about really that auditing and keeping your. Your language models up to date.

[00:35:06] Isaac Sacolick: I love this comment here from Vaibhav on the comment stream. He asks it as a question, is it a good idea to identify all legacy applications and take modernization project for the same on top priority? I’m going to switch his words around. I’m going to say it is a good idea to identify yes, this is exactly what we want to be doing. And I love this idea of a lineage or an audit log of learning.

One of the things I’m always concerned about when we talk about AI is the data you’re feeding dated and does it know how to think about what’s new and important versus what’s aged and outdated and really important. When you think about operations, you think about field and you think about construction, where so much of the type of work that we’re doing is changing all the time. Heather, welcome to the floor.

You’re our people person.

Lots of things are changing around us and we’re trying to create a culture where people are simplifying and using technology and using AI to do that. I just want to. Where are your thoughts going on this today?

[00:36:16] Heather May: Well, a couple of things and one thing that I want, a lot that I agreed with, one thing I’m going to take issue with is that it’s not about an age Thing I do think it’s a generational thing that the way people that have been doing work for 25 years, it’s generational, they’ve been doing it that way for so long that they don’t want to share it. And I think that concept of being afraid is brilliant.

How do you counteract that is you celebrate them, you bring them together, make it comfortable, give them a psychological safe space where they can talk about those stories and you capture those stories. And from that you can take document things. And certainly the technology is there to do the documentation. The process is there for the documentation, but you still have to collect the information. And so much of it will happen in their war stories, in highlighting their successes and their failures and being able to quantify and identify where those different steps are. So I think that. And the newer, younger generation will be able to do that. The technology is very familiar to them. It’s not familiar to the older person. But they can talk and they can share those stories. And if they’re given that opportunity, it can be there. And having them realize that this is just the first step. Because so much of what knowledge My background has been in research and as soon as something’s written, it’s old. So that documenting the ongoing and yes, and auditing, all that’s important. But it has to be in real time. It has to be all the time. It has to be a dashboard that’s shoved in your face that says, today, this is what happened yesterday, this is what happened. What can we do to fix this?

So I think that has to be really emphasized. And it is a question of a cultural shift wherever it’s going to start.

And it can’t be just in one place.

It has to be a pervasive feeling that this company, whatever company it is and whatever industry it is that you’re being able to, to say to people, we are changing, you are part of this change. You’re not going to lose your job. Don’t be afraid.

Let’s move this forward together.

And once that happens, and then you can deal with those succession plans and you can deal with the real time support and you can deal with the compliance so things aren’t left to chance. I saw an article that just said you can’t leave it to betting on the way things are going to be. And you’re right, you can’t.

[00:38:52] Isaac Sacolick: Heather, I agree with you. I do like the word generational issue, not an age thing.

Because when you look at the younger generation, we tend to use a different word. We call it a Hero culture.

The person who has figured out, has developed something, has created something and left it in such a state that she or he is the only person who can come in and fix.

Goes back to Martin’s story.

Who’s being brought in and how are they actually fixing things? They’re not fixing things in a way that others can contribute. And I think that’s all about not necessarily documenting things the way we think of in the past. Right. I still have books in my attic of the documentation guide for software that I created. I use it as like a memoir. We don’t need 400 page documents. What we do need is people to log their decision for people to share an insight to teach others. Derek, you’re in the security space. I mean like knowing how to solve a particular problem when it comes up again is not about creating a massive folder of operating procedures and security.

[00:40:06] Derek Butts No, you’re right. And the part of the problem is understand that you have a problem. I think that’s the first thing to look at. But just to kind of go back and then looking at the challenge you mentioned before, you know, when you’re talking about the risk areas, the two day I see a third party risk and supply chain cybersecurity risk. These are some of the most prevalent ways people have been breached because of lack of consistency going across the board, just, just accessing who has access to the data. And when you look at this and bring it forward and you talk about addressing these gaps, you know, we go back even further, we talk about the documentation. Well, artificial intelligence and the innovation. Artificial intelligence is only going to take that data that’s been properly marked and accessible and make it audible. You can make it automated. And if you don’t have good data, you’re not going to have access to data. Because you know, with artificial intelligence you’re going to have to train these systems to work with these automated tools to pull that documentation. I also look at things working with, you know, just the, the automated tools you can need to have in place now for the tribal knowledge and how do you close the gap? Having these tools that manage the threat associated with data and the tribal knowledge is being lost or building these dispersed systems, how to pull them all together so they can be unified.

These are things that we need to be happening at an enterprise level. And those management, management has to identify I’ve got a problem, I need to fix this. How am I going to do that? What systems do I need to put in place? I’ll kind of look at these knowledge gaps and graphs to show who’s doing what, what are dependencies, how do I visualize this? As mentioned by Heather, with a dashboard of some sort to show these are the things I have, these are the risk associated with the things I don’t have. How do I bring it all together? So it really is going to be looking at how to pull it all together, how I’m going to manage it from a governance and a policy point of view. You know, all these things come to remind you. Now you mentioned about, you know, the documentation and the amount of documentation. What is considered too much and I agree with you, 400 page documents or anything is going to be too much. But if I can automate it and assign accountability to it and assign roles and access to it based on certain individuals, it’s going to make it a whole lot easier to manage across the ecosystem.

[00:42:12] Isaac Sacolick: Yeah, look, I think this is about giving people tools where they can share snippets. You know, what did you learn today? How did you solve that problem?

What went into deploying this feature?

What was the change that you did in the operation to make this work better?

I think about all these. I think of it as a small problem of sharing that knowledge. And I’m looking at AI to my rescue, of being able to look at all the contributions people are making into these knowledge databases and saying, you know, now help me discern this into a storyline, help me query it. This goes back to something Liz had said earlier.

Martin, where are we going with this?

I mean, the reason I asked this is because I just do feel there’s a moat there. You know, going back to what Bob said, that if we don’t find a way to have really strong, real time, up to date information about what we’re doing in our operations, it’s going to really be hard to put AI in place now that it’s becoming more accessible and at some point more affordable for industries like manufacturing construction to take advantage of it.

[00:43:30] Martin Davis: So I’m going to take you to the manufacturing floor and this is kind of building on an example I’ve used in the past.

So you’ve got a maintenance engineer in a manufacturing plant and they’re walking the line and they have a head up display, an augmented reality display that is showing them information about the equipment they’re looking at.

And that information is coming from lots of different sources. It’s coming from sop. So I’ll go to where Bob was going with sops, small language models of all the data we have, manufacturing information coming from sensors on the actual equipment. It’s coming from the actual equipment manufacturer themselves, providing maintenance routines and other data like that.

So person can be looking at a piece of equipment and seeing how it’s operating.

Let’s say they need to do some maintenance work on that piece of equipment they’re looking at and be able to bring up the previous maintenance history of that equipment. They’re able to bring up and show them on, on the head up display. They’re wearing the actual maintenance steps. So the sop, the steps to actually change, change a bearing or whatever it might be, or they’re actually able to then. So we’re taking it from the data that’s captured from various sources being brought together using the latest technology, using a small language model, using AI, using your sops, using the data to help the person actually conduct the task. So it’s funneling everything down to just what the person needs to do.

[00:45:11] Isaac Sacolick: Just kind of.

[00:45:12] Martin Davis: I sometimes think it’s very good to make some of these things real in terms of this is actually how it can work.

[00:45:19] Isaac Sacolick: I love that picture. I even love it in multiple use cases. Other than just maintenance. I’m thinking about putting that AR display on somebody who’s learning what equipment is on the shop floor and how it works and what the different functions it does. I think it’s a learning tool. I think it can be a management tool as well to see how things are performing and where things are having quality or performance issues. I think there’s a lot we can do as we’re using tools to collect data and then using AI and using display technologies to bring it right to the people who are working on the floors or in the field. Hey John.

[00:46:02] John Patrick Luethe: Yeah, Isaac. Yeah, I, I love what I’m hearing and I, I agree. If you want to have better documentation, you got to improve the tools to make it easier for people to submit information in blog tickets on the documentation. I was talking to a guy that he works for one of the large car companies and I saw his jacket and I asked him about it and he said he loves that company because management listens to people. And he said that every time he finds an issue with the documentation or he finds a better way to fix something, he writes it up and if it gets accepted, he gets 300 to $1,000. And so that’s like such an amazing thing for him that people actually listen to him and then he’s so happy to get the money.

The other thing that happens is that if you can put your steps, put your repair guides in electronic format, take a look at what people are looking, you can look to see if people have recently looked at that thing. We were at a company, we found that it’s really hard to diagnose things, but once you know you diagnose it, it’s easy to fix because the repair guides are real good. We were keeping track of which steps did people do and what was the success or failure of those. And we were able to look at and get the. Basically by doing machine learning on the repair guides and the steps that people were taking and the success or failure of those, we were able to really improve the repair guides and serve up kind of dynamically what people should be doing for diagnostic steps. And so if you invest in better tools, the documentation gets better. And if you incent people and treat people well and listen to them, the stuff gets better too. And that improves culture and, and then they actually, you know, if management listens to them, a lot of good things happen.

[00:47:36] Isaac Sacolick: I think you talk about better tools and incentives. I love this $300 to $1,000 bonus for recommending process changes.

Joanne, you’ve worked in this space, you have a startup in this space and it’s all about process and data and culture of sharing. Where do you want to leave us today, Joanne?

[00:47:57] Joanne Friedman: Where I want to leave you is with two points. If I take Martin’s example that he was just relating about the heads up display. The one thing that I did not hear from him was the inclusion of the individual’s knowledge as well, the tribal knowledge that was the piece that was missing. And you know, when I look at these things, one of the things that we have and are working on is agentic digital twins. In other words, taking a digital twin, which is basically a mirror of what a piece of equipment is on the factory floor or the entire process. And then adding to that the notion that you can now take and go from what if meeting simulation all the way to what’s next as you’ve put all of this data together, the inclusion of the tribal knowledge and the ability to inculcate the value that John was talking about as well, which is, as I said at the outset, incentivizing people, you giving them a different way to learn and to teach meaning, to impart their knowledge. To Heather’s point, this is all the capability that can be built into a gentic AI because the agents can be differentiated by the task. You know, take the big process, break it down into smaller chunks, etc. Etc. We’ve all heard about that, but it’s what you do with each of those chunks that makes the difference on how productive the agentic system may be. In the case of a digital twin, you can actually take a visual image to what Martin was speaking of, see exactly where the problem is, see what’s happening with it. I’m noticing the same thing is happening. You know, my son is a master electrician. He’s getting a lot of innovation in his trade where he can now use some of the more sophisticated tools on the market. Not only to help him get to root cause analysis faster, but he can add his own perspective. Like, listen, I’ve run through, you know, thousands of lines of code books, I’ve run through manufacturers specifications.

X does not work, Y is the fix because there’s been an error in the documentation, or there’s actually a better way to do it. Faster, cheaper, you know, more successful in a particular situation.

Those kinds of nuanced information are what is absolutely required going forward. How we, how we institutionalize new knowledge, not standard operating procedure, but the changes of the standard operation needs to change as well. And, and the tools that we give people, whether it’s, you know, I know one company that we’re working with where they’re using 30 second video clips on TikTok.

What, what did you fix today? What was the problem? How did you solve it? Impart your knowledge and they have a contest going on where that’s how they’re using new technology like, you know, video stories or storification to get the knowledge out of their workforce. And the younger people may be more inclined to do that. The older people may want to sort of have it as a conversation where they’re not being filmed, audio, video. We don’t need the physical piece of paper mentality or capability that we’ve had for too many generations because the original standard operating procedure is no longer the same. The documentation issue, whether you’re using an AI tool to do it or not, is too much.

And you don’t necessarily need to know everything that’s in that documentation to get to a root cause analysis and a fix. Time is what’s important here.

So the more we can impart that through, whether it’s agentic AI or another form of AI physics, AI, for example, or math, we’re ahead of the game. And those companies that start adopting that capability, they’re the ones who are going to lead because they’ve taken into account that tribal knowledge or institutional knowledge is actually one of the most valuable assets of a corporation. It’s time to treat that data as an asset.

[00:52:27] Isaac Sacolick: Well said, Joanne. Joanne, you and Bob both mentioned standard operating procedures and I think what we’re really talking about here is dynamic work management. It’s really absolutely.

Quickbase has a lot of writing on this on their website, I think about weather factors, supply chain factors, people who should be in the office or at the field today who aren’t there because of a personal issue. Right. We tend to think standardization means a nice straight line from point A to point Z and a finite number of exceptions that we can manage to. What I think about today’s world is lots of things can go wrong, lots of things can surprise us today.

And where our subject matter expertise comes in and the antithesis to this is tribal knowledge is what to do when there’s an ex, when there’s something you didn’t expect.

How do we.

[00:53:32] Joanne Friedman: Well, this is.

[00:53:33] Isaac Sacolick: Go ahead.

[00:53:34] Joanne Friedman: Sorry, sorry for interrupting you, Isaac. This is where, you know, newer tools like knowledge graphs start to paint a much better picture. And I would say, you know, as an example, something that we’re working with at the moment is a business process that is multivariant, has all of the kind of complexity that you’re discussing. Weather factors, you know, closed loop supply chains, open loop supply chains, remanufactured goods, greening, all of these capabilities coming together.

Knowledge graphs is a way to paint that picture for people in a visual representation where they understand the what we couldn’t understand from a standard operating procedure, which is not only the nuances but also the dependencies and codependencies.

And this is where I think all of you know, change that we’re about to see, not only driven by AI, but just generally speaking as a result of AI being part of our world now is going to come into play. We’re going to be much more aware of those cross dependencies, ways to reach, you know, through silos and, and, and have a bigger picture to deal with.

[00:54:55] Isaac Sacolick: Thank you, Joanne. Great, great comment stream happening here on LinkedIn Live. If you haven’t said hello on the common stream, do drop in and say hello to everybody. Want to thank Steve and David, a whole bunch of people who’ve been sharing their insights, some that I have put on the whiteboard that you see here.

Joe, love to hear your thoughts on how we close the gap with tech and AI and then we’ll bring Bob on for some final recommendations. Hello Joe.

[00:55:23] Joe Puglisi: I think Bob, Martin and Joanne have all said very eloquently what I would say is documentation is not static. It’s not a point in time, it’s not a task, it’s an Ongoing effort with a continuous feedback loop. Like many other systems in the construction industry, electricians use a lockout tagout mechanism. Right. And it’s, you know, pilots use takeoff checklists, but those, those things are static.

And when we can bring technology to bear on delivering that kind of knowledge and expertise about how to protect yourself and protect your customers, your passengers and whatever it may be, when we can bring that in with real time feedback, as Martin and Joanna both suggested, that’s when you start to really win.

The companies won’t win that have the best knowledge management systems.

They’ll be the ones that make tribal knowledge obsolete by making expertise ubiquitous.

[00:56:32] Isaac Sacolick: Making expertise ubiquitous. Wow.

What a, what a leading statement for the age of AI where we need to be able to ask the right questions, where we are going to use AR to bring more, more real time relevant information in the field where sharing is more important than doing at times.

Bob, how do we get there?

[00:57:00] Bob Salaj: Wow. I think we’re all going to have to come back and rewatch this entire episode because of how many good points are kind of made across the board. And I think the real cost of tribal knowledge, right, I think really what we’ve been talking about and there’s a lot of impacts it to me, I start thinking about the intersection of like workflow, person and data. We talked about the culture, we talked about the operations and the tech side where AI is going and really how do we kind of move past the static ness of really it’s taking too much time. And I’ve been watching the comments and tremendous comments there and I think really at Quickbase we can imagine a world of AI not only for building applications that really match your workflows exactly and mining the data, but it should also be documenting itself so that you don’t have to pull out a number two lead pencil and actually ask someone because of that change of culture of actually documenting is so tenuous and tedious to kind of move forward with the system and AI should all be doing that all together so that the impact isn’t taking 17 months to ramp somebody up or losing three months. I think in terms of the. If people leave then that’s going to be the impact to your process. So how do we move all that together into one Venn diagram so that we can help out the output. So I think it’s exciting times ahead.

[00:58:19] Isaac Sacolick: I thank you for that, Bob. I think most of the folks here know that I’m a big fan of Quickbase. I use it to run my own business.

Every one of our coffee hour episodes. Every one of the articles that I write ends up in a Quickbase database. It’s how I manage my business.

And you know, increasingly what I’ve been doing more with it is using it as a knowledge tool, using it with AI and to just be able to ask questions around simple things. What should I write about next? What should be our next coffee hours?

And you can’t do that unless you have centralized knowledge, unless you have things really typed with strong metadata and at times being able to permission access to the information so the right people have access to it. So a lot of really good knowledge here. You mentioned we should go back and relist listen to this episode. This episode will be on my blog drive.starcio.com Coffee.

It will also be up on Apple and Spotify podcasts. Be about a week before we get it up there. This dashboard will be cleaned up and you will see a blog post with me from me on Monday with some of the best practices here. So lots of really good knowledge here. Thank you everybody who’s been participating on the Common Stream. My speakers here. Bob, thank you for joining us from Quickbase.

Quickbase is what they say to me. Say goodbye to workarounds and hello to workflows that work with Quickbase. Learn more@Quickbase.com Solutions if you ever want to see a demo of Quickbase I have loads of them.

Happy to show anybody who is interested in what it looks like in my world. Obviously I don’t work in the manufacturing or construction industry. We but I’ve used those examples in some of my previous episodes. We talked about SAS sprawl before and lots of opportunities to use Quickbase to use to cement out field operations. Folks, thank you for joining this week. If you look in the top right hand corner I have the upcoming episodes for October up listening there in our whiteboard. You can also see them on Drive Star CI.

Thank you for participating this week everybody. Have a great weekend. And again thank you Quickbase for sponsoring this episode and thank you Bob for being our expert on this particular episode. Everybody have a great weekend.

[01:00:56] Bob Salaj: Thank you everybody.

This episode is brought to you by Quickbase.

The views and opinions expressed herein are those of the speakers and do not necessarily represent the views and opinions of Quickbase.