Your Safety Metrics are Lying (ft. COMET)

Hello everybody. Welcome back to the podcast. I’m Monica and I’m joined by my co-host Scott. And today we’re really

excited to be joined by Mark Rushton, CEO of Comet. Mark brings a really

interesting perspective on how organizations are thinking about data. Not just what they’re measuring,

but how they’re actually using that data to make better decisions. And before we jump in, just to level set for those

folks who are not familiar with any of this stuff, which I’m throwing myself in there. When we talk about EHSQ or QHSC

data, depending on what side of the pond you’re on, we’re talking about the data organizations collect around quality,

health, safety, and environmental performance. So, think of things like your incidents, your audits, your

inspections, and then the near misses. So the reality is that most organizations actually do have a lot of

this data, but the bigger question is what they’re really doing with it. So in today’s episode, we’re going to be

zooming out a little bit and talking about how HSQ data has evolved, why

expectations from leadership are changing, and where there’s still a gap between having data and actually having

real value from it. Hello, Mark. It is so good to have you

on the pod today. Thank you for joining us. Uh do you want to just take a moment to introduce yourself and Scott as well?

Feel free to jump in at any point. Fantastic to meet you, Scott. Great to great to see you again. Uh yeah,

delighted to be here. Mark Rush, CE CEO at Comet and and for a long time now. uh uh someone who has been intensely

curious about what organizations are are doing with all that stuff that they’ve been gathering for years and years and

uh and often organizing and storing very well in in in wonderful systems like intellects and uh and now are becoming

more increasingly curious themselves about how they might unlock some of that value and some of that hidden insight

that might just be lurking inside it. So looking forward to the debate this afternoon.

Yeah. And I’m Scott Gatis. I am vice president of safety and health uh here at Intellects and I’ve been in uh the

practice of EHS for nearly four decades and things are changing uh around me

daily it seems. So I’m looking forward to uh to this podcast. Lovely. Me too. So the first topic that

I wanted to jump into was around data explosion. Um so Mark start at a high

level. QHSC data is more accessible now than ever. what has changed over the

last few years to get us here? Uh, great question and and you’re absolutely right. It explosion is

probably the right word. Um, and it’s not an explosion born out of just new technology. We’ve been gathering and and

uh and storing and organizing this data for a an awful long time now. five ten

years of of rich data is not uncommon in most semi- mature uh organizations that

take their their quality their health and their and their safety seriously. Um you know in your world many of them

because organizations that buy intellects already have made a decision to take this stuff seriously. I think

the big explosions come in the ability of uh I suppose technology to help unlock what’s what’s what’s hiding

inside all of those uh all of those repositories and not just to unlock it but to put it put it in in in in order

and understand you know how to how to mine for some of the very specific things that that data might might be be

holding that might give organizations the the chance to be a little bit more decision centric uh based on the

strength of signal that their data is giving them about maybe something they need to invest in or perhaps something that they don’t. So, I think that the

the explosions come not just in the fact that we now have more data because we’ve been collecting lots of it for a long

time. It’s much more around now having the ability to actually to unlock what’s held inside it.

Yeah, I I would add to that. So, Mark, I I totally agree with you. I I think when I started practice nearly 40 years ago,

we were just okay with data regardless to where it came from, how it was presented,

right? And much of it was lagging data and and we were satisfied with that. I think now the real fight today is making

sense of all of this data that we have. And I can look back at intellects probably 10 years ago and said we had

most of our customers had just simply wanted to collect data, right? And and

what I noticed uh when I was leading companies um for environmental health and safety is that leadership right has

started to notice what datadriven decisionm looks like and they’ve learned it from finance and from operations and

supply chain. Now they’re expecting EHS to do much of the same, right? They’re expecting EHS to play at that that same

level. And and I don’t I don’t think that’s a threat. I think is an invitation, but the practitioners really

need to step up, right? Really need to really understand that we have a place at that table if we want it, but we’re

going to have to think uh maybe a little bit more businessminded. And I think that that’s where technology really

helps uh because it fundamentally changed not only the accessibility to data, but we’ve been able to pull it out

of spreadsheets and siloed system and now we can centralize it, we can structure it, we can report it in real

time. So the accessibility right has really made things happen differently I

can even say in the last two or three years than than the the previous 20 you know years

100% and and and your example around the comparison to operations I think is really really important if you think in

the world of ERP and and and operations typically the currency within those

systems is cash those things are directly money related they’re related to efficiency and profitability and and

and oper operations and project delivery. The the the currency inside an

EHS management system is data. That’s its currency. And so why can’t we unlock it in the same way as we unlock the cash

savings that can be that can be retrieved from from well-managed operations and operations technology. So

I completely agree. I completely agree. And of course we now have this thing in our world, don’t we, that’s very mature

called data science. you know, the ability to to to create models and organize the the the data in such a way

that it can be actually pointed towards giving you things that you couldn’t get before. So, you’re right, there’s there’s a huge opportunity to to to

unlock all of that effort of the past decade in particular and really start to to use it to make to make decisions. And

uh I’ve got this kind of saying we have in our world that, you know, counting, but does it count? And and I think I

think that’s a really interesting way of of of framing some of this stuff up because for me numbers don’t really cut

it anymore. I don’t really look at numerical data now and think, “Wow, that’s telling me something really earthshattering that I

can go and make a decision on.” Okay, I’ve had 10 of them and 15 of them, but so what? I want the color. You know, I

want the detail. I want the descriptors. I want to understand the context so I can make a decision. And I think that’s

where where you know the the the technology is becoming much much more of an enabler.

Agreed. Agreed. So with all this data now available though, how do you think expectations from leadership have

changed when it comes to this performance? Would you say the leaders are asking better questions? Would you

say they’re overwhelmed? They don’t really know what to do with the data. What do you think? I think there’s two camps. Um, I think

there are there are still leaders uh out there that expect you can just throw this stuff over the fence and you’ll get

better answers. Um, and they they they’re still on that kind of learning uh journey to understand what needs to

be right in order to get the right kind of insights. But the switched on ones are are definitely now being more

demanding of the um exploitation of all that effort that’s gone on in the past.

And quite rightly so. um you know we should be as as EHS leaders much much

more confident about making more informed decisions if we use the data at our disposal more effectively. Uh the

the fact is you know that that that every organization in the world that we’ve ever come across and I’m sure

you’ve ever come across will have a a smaller percentage of its bad actors that give it disproportionately more

pain than all the others. And so using using this type of analytics to identify

and isolate what those things are and understand their behaviors and their habits and and do do more work to

mitigate them will will definitely result in in more sustained performance uplift. So I think that the informed

ones are quite rightly much much more demanding of of what their expectations around this this genre should be and I

think quite rightly so. Yeah, I I don’t disagree with that either. And I I I predicted that we

would not disagree. uh much on anything you know for for this podcast because I

always look at things as a practitioner first and I have sat through I don’t know how many executive reviews with

boards on senior leadership teams and the headline number is always one thing

it’s an injury rate and if it was down everyone excelled and and they moved on

right a low lagging indicator it meant something different to that group than

it did to me what What it told me is that well we were measuring things that had already happened in the system and

building strategy for lagging data alone was always a look back into the system

and and I would say I would preface that by saying I think it’s always important to look back you know I’m all right with

that but over reliance and presenting information to an executive team that does not grow them in a different way I

think it just really hampers us as as what the practice of of eh s requires us

to do today or even quality. Uh so it it has been one that uh it felt like

pushing a rope you know at times but I think it’s changing and I I think it is I think it kind of goes back to my my

first comment is that we’ve learned a lot through operational effectiveness through how finance is reporting you

know their information now. So, so we’re seeing strategy change and and we’re even seeing you know our own clients

really ask right to help them have a more balanced measurement portfolio so

they can have conversations with executive leadership teams. So the goal is I think for us always forward looking

of where risk is actually building and some of that is looking backwards but

much of it is actually looking forward into the system. Yeah, I I agree. I violently agree and

and uh and and that’s interesting and and let me let me pick on your your your core profession a little bit for a

second Scott because I this is maybe a little bit unfair but historically EHS professionals have been very good at

beating themselves up about the things that don’t go right but but to your to your point haven’t spent nearly as much time in

understanding the the signs of things that do go right and my my kind of epiphany in this field was about five

years ago now I got exposed to a an assurance insurance data set. It was a data set around it was about 28,000

interventions across a number of years across a large energy company and and and anyway analysis of this of this data

showed compliance rates in excess of 95% and then and then even where there was

non-compliance the sliver of activity that was more extreme non-compliance was only about 10% of the 5%. So it was

really really and the first thing I turned around and said was have you done any analysis of

all these times you get it right and of course this is now called and and and and all the other things

that come with it but back then it was like well no there’s an assumption we’re just good at what we do and you think well no but there’s there’s things in

here that are telling you why you’re good at what you do let’s spend a little bit more time looking at that and and that spawned a a really exciting kind of

kind of uh caveat within the technology now which I really like which is aentic AI where where you can you can train

agents to look for very specific things like compliance to a particular standard and that can be worth its weight and

gold because it can tell an organization what they’re doing right and how to do more of it. In other words, get on the front foot.

To Scott’s point, use those leading to to reinforce the good. There was a a

project we got involved in with with a a research body a couple of years ago where we looked at compliance to a rail

standard called uh called RM3, which is a a risk compliance score for the rail

industry. And it was incredible because when you applied a data set to it, you could train that that that agent to go

look for compliance in that specific standard. And that’s great because it identifies gaps before they become

consequential, but it also it also allows you then to get on the front foot and look at what’s really really good.

So, you’re right, Scott. There are there are there there are lots and lots of ways we should be looking at getting

getting ahead of this before we’re looking back at maybe a a statistic which is quite a blunt instrument to be

to be frank and and over the piece doesn’t change massively if we look at serious injuries in particular.

So, uh yeah, I’m totally in that camp. Agreed. Um, well, in the spirit of

talking about indicators, we did talk about leading indicators, but I also wanted to talk about lagging indicators,

which is still something that dominates. Um, why do you think that organizations

still rely so heavily on these? Is it because they’re easy to report? What gets missed? Um, any insight do you

have? Uh, yikes. How how political do you want me to be?

There’s there’s a bunch there’s a bunch of reasons around kind of TRI and and and and lagging indicators and and some

of it, you know, is is is slightly worrying in that it’s not uncommon still

to find industries where where bonuses, etc. are are are delivered based on on

that that particular um statistic. And as a result of that, you get things that aren’t reported properly. you get you

get LTI’s reduced in RWC’s and and you get laptops being sent into hospital wards and all sorts of things, but

luckily that’s changing. There’s not so much of that anymore. I think it’s an easy an easy to an easy statistic to

take a view on. And I think sometimes those that aren’t involved in the profession, maybe CFOs, CEOs, COOs, they

maybe can understand that quite quickly. And to Scott’s point earlier, when it goes up, they know it’s bad and when they go down, they perceive it to be

good. Um but but but there is so much more color to be unlocked from looking

at looking at a far broader um uh portfolio of information that QHSSE or

EHSQ provides. And and I think that’s that’s maybe where the where the where the smarter money is is starting to go.

It’s just the simplicity of it. I think the bluntness of it almost is one of its appeals. Um but but but yeah, we’re

finding now that that that less and less switched on organizations are accepting that as a as a as as a primary metric,

not least because it doesn’t tell you very much. Um which is uh which is which is ultimately the key. It’s not the the

number that matters, it’s the why that sits behind the number that really matters. And if we can get into metrics

that unlock that why, then uh that’s much more interesting. Yeah, I I I totally agree with that. I

think that if I look back, this started in the 1970s when OSHA produced what

they called the OSHA recordable, right? And they defined it and uh and then they had senior leaders started they started

understanding what that number meant or what they thought it meant. Uh and they have literally been trained decade over

decade over decade that that was the measurement of success for an organization. And it’s been very

difficult to turn back time, right, for for practitioners talking with a leadership group that has

fervently defended that number because it seems easy, but it doesn’t tell us enough. And it’s exactly what what Mark

said. It’s it’s not the why. What there’s no why behind that number unless you’re looking back at at failure. And

when I think about that, I and Mark, this this kind of spawned a thought when you were talking on your last comment

about the time I I’m in front of the board. I’m talking about our reportable incident rate. they they’re they’re

thumbs up or their thumbs down. And I was talking about a specific facility that had a it it was a good total

incident rate, but what their workers compensation what their workplace medical costs demonstrated that they

were by far the most solid member, the most solid facility that I had in the entire in the entire corporation. So

they were controlling uh not only right recordable incidents but they were controlling medical cost which to me is

many times how I define if I’m looking at trailing indicators how I define success with with a facility but it was

a difficult conversation right it was very difficult to have a conversation outside of that one siloed metric that

they were locked in on and and that’s what what we have to overcome Monica when we talk about this is that uh an

over reliance regardless So what that lagging indicator is, it’s an industry-wide problem and it’s one that,

you know, with folks like us at at Common and Intellects, we’re trying to break that. We’re trying to have that

conversation that it is more balanced. It’s it and and I’m okay with understanding lagging indicators. I

think it’s important. I think I’m very okay with understanding those leading indicators and blending that story

together so you can have the conversation. And uh and for me that’s the most important part that I have in

all of this is to be able to have an intelligent conversation about where we need to go.

I think what I’m hearing and keep me honest is in order to have those intelligent conversations, you’re going

to have to take one step back and have the dashboards and the insights that actually provide you with that

information. So do you think that these dashboards typically fall short? And what does good actually kind of look

like here? Yeah, that’s a really interesting uh that’s a really interesting question. I

mean, the the beauty of of how you how you report what it is you seek to report is that that’s very simple now. The

visual interpretation of what people see and read now can be can be can be very very compelling. So, it’s unusual to

find to find um you know well- constructed dashboards that don’t tell a better story than maybe than maybe maybe

what was what was previously possible. I think I think you’re you’re right on the on the kind of step back thing. What is

it what is it that’s going to add value and how do we need to consume that information to be on the front foot and

there’s there’s a couple of there’s a couple of things I suppose come into play here. One is around at what point

does data become actionable? At what point are we confident enough in it that we’re going to be able to go and make a

better decision because of it? And that’s often not a decision about spending money or time or resource. It’s

sometimes helping you make a decision about where not to spend money, time, and resource. I think the overarching

the overarching theme behind this discipline is that there’s not an organization in the world that has an infinite budget to keep itself and its

people out of harm’s way. It’s got to know where to put things and where not to put things. And I’m going to bring

the P word in a little bit here, but I’m going to use a small P, but prediction is of course the the kind of holy grail

of of of a metricled output. Can we get to a point where we can predict when our

next failure or where our highest risk lies? And I’ve always felt the kind of tools that talk about telling you that

you’re going to have an incident in stairwell 13 on 2 p.m. on Friday in damp

conditions is dangerous. I don’t think that that level of prediction is is necessarily possible or healthy. But I

do think that if you get this stuff right and you use those leading indicators properly and you challenge

yourself beyond looking for numbers then prediction with a small P should become very possible where we can start to say

well actually that that the systemic risk is lying here. This is the particular type of operation. this is a

particular type of client or project or or location or geography whatever it might be where we’ve got a bigger risk

and that’s born out of all the the lagging and leading indicica indicators that we’ve got at our disposable but

typically that type of insight doesn’t come from looking at numbers it comes from looking at all the things that are

driving the numbers so if for example we’re looking at what barriers within our safe system of work were effective

or failed or missing it’s no good to tell me I’ve got 10 five and four I need to know what they were. I need to know

where they were. I need to know the conditions around them. I need to know potentially if there was human error associated with them and what that human

error was was was telling me. All those things become important and and really the the role in technology here is to

bring all that to the four and organize it and structure it in such a way as we can we can interpret the results quickly

and and and and clearly. And that’s still not easy because in large organizations, we’re talking about

literally hundreds of thousands of records and getting those those real insights within those records to a point

that we can we can have that color that we need. Yeah. I I think and it’s one of the the

uh the wonderful things about our relationship with with comet is is when I look at data right presented in front

of me whether it’s leading or lagging I think genuine insight requires

interrogating that data point right we got to be willing to say well that

number looks fine but what’s hiding up underneath that that number and you know

I’m writing right now about how senior leaders should act in the workplace and it’s a lot around around this topic is

that I find a lot of of senior leaders they’ll look at a dashboard red yellow green you know full of needle gauges and

they won’t respond to green and that’s a that’s kind of a lean mechanism right is don’t respond to green only respond to

what you see red or yellow and those are the things that you start you know questioning I I would challenge that I I

want you to also question why am I seeing green what’s changed in our system and of course then challenge you

know yellow and and and and red as you see those things but I I think that the

most benefit that we have as a leadership team as practitioners is to interrogate the data that we’re seeing

in front of us and comet you know this is your your bread and butter right is that we’ve got causal factors we got

root causes we got system features that all need to be talked about and that comes through I think you know through

conversation and through development of a skill that we we just don’t have today. So that’s a practitioner skill

and I think it’s one that our profession really needs to own very very aggressively. I think Mark says

something because I also use prediction with a little p. You know, in my mind, it’s always a little P. And it’s very

difficult for me to even talk about prediction and what I do as a practitioner because uh the biggest

variable I believe in the workplace is a person, right? And trying to track that

person and exactly what that person is going to do in the environment that they’re going to be doing it in is very

difficult. I I look at it almost more like prescription. Can you give me two or three things that get me down into

this funnel of things that are probable? Right. But I I’m in the seat as a professional, right? So I I’ve got a

gut. I I can see things with my eyes in front of me and things change. You know, I think Mark, it’s exactly why you said

weather changes. It’s too dark, right? It’s it’s too light. All of those things do matter. So helping me get to an

answer is more important to me right now than giving me that answer that I’m going to maybe question a little bit

more. And and I I think that there is a shift happening. You know, there is a shift of giving us better insight that

is real and meaningful. I think dashboards answer the what in many cases, but we really have to be a part

of the why to understand what’s next, you know, in the work system.

I think you bring up a good point about the the small P. Do you guys think that we should be utilizing AI to help us do

better predictions and help prevent some of this and provide that insight that data sometimes can overlook

one 100%. Um it’s it’s it’s not a case in most forward thinking businesses now

about if it’s a question of when if not now. Um and and and we’re fortunate in this profession actually because EHS uh

data typically is benign and and and it’s stuff that we can we can benefit as industries from sharing and we can

collaborate on and and we can do we can do good stuff for society. We can do not for profofit stuff around that kind of

data because it’s about keeping people out of harm’s way. If you tried to do this in the realms of competitive data

involving IP and commercial, you know, stuff in organizations, we’re years away from that in in a collaborative sense

because it literally is so so difficult for organizations to protect what’s theirs. So, we have an opportunity in

this space to use AI to unlock a lot of what we what what we’ve maybe historically not been able to see. And

the use cases are varied, Monica. So, you know, I I I call one of my favorite

use cases daily intelligence. If you’re nosy and you want to to see changes and things that may be evolving in your

data, it’s a great way of going in and and using it as your as your newspaper. In many ways, you can use it to satisfy

a hunch, you maybe think something’s wrong, but you don’t know. And you go in there and you and you have a look. I

remember being involved in a case recently with AI in the rail industry where a a small child wearing Croc

sandals got their their the croc ingested in the top of an escalator and they lost two toes and immediately that

whole network rail operator was able to go into their data and isolate escalators and look to see if this was

something that was maybe occurring or nearly occurring elsewhere. So those use cases, those hunches can be met much

more easily. Sometimes it’s because you’re planning to deploy resource over a year and you want to know where best

to get the biggest value from that resource. That’s really significant. I think that’s a really important reason

why AI needs to be needs to be an important uh an important tool within your your EHS management toolkit.

Sometimes it’s because you want to repeat as Scott said what’s going well. You want to look into that green and uh

and that’s great because that has a direct impact on systems like intellects because those preventive actions that

you are responsible for managing the roll out and implementation of they should be instigating bigger green and

so understanding of the effectiveness of those actions through using AI le insights is is really important as well.

But my one note of caution is us humans we have to retrain our brains to be the QA. We must be very careful about using

AI as a decision-making tool entirely in its own in its own right. It’s a

strength of signal. Sometimes that strength of signal will be pushing through the roof. It will be so strong and there’ll be very little risk in the

decision that you’re going to make, but sometimes it’ll still be in that area where your human judgment and your human

expertise is absolutely critical to making sure you get that right. And I

pick on North America a little bit sometimes, but it’ll only be a matter of months before there is a huge court case

somewhere because someone taken AI in a tool and used it verbatim and not had

the outcome they’ve expected from the million dollar investment that they’ve made. And we don’t want that to happen

in EHS. We want that the value of the AI to be understood, but the value of the human alongside the AI to be understood

equally well. I hope that sounds fair because I think that’s that that’s probably where we need to be. I I I

think it’s completely fair. I I you know, this is one of those those areas that I kind of dive into and I I I guess

I use AI every single day, not just with with the platform, Monica, but actually

generally thinking about things. Uh most of the time I I use AI, right? I’m just

trying to validate some of my own thoughts around a certain subject and I’m always amazed that it comes up with

those two or three things that I didn’t think about, right? or maybe I should think about or maybe something I can

disagree about. So for me it’s a challenging exercise to validate what I

really want to say or what I want to write. So I think it’s good on those. I I speak about this a lot. You know, I

teach a u a session at many conferences and it’s one abstract that’s been

accepted widely is uh chat g our our chat hacks for the safety professional

and it’s literally just a session on prompts is how do I write the right prompt to get the information that I

that I want to get. That class fills up. If it’s a room full of 500, 500 people will go into that room because they’re

always very inquisitive about how to use AI much better. And I think that’s a good thing. Uh I I do like having access

to large data sets. I like that that ability to do that. So uh when I when I

compare it with what we’re doing with intellects, um you know, we’re certainly using it differently and I set on that

AI team and the COE actually does that at and we just watch, right? We we sit

back and and that team brings forward an idea. We look at it. we’d vet that idea

and many of them are about 70 to 80% there when we see them and then they

work on that next 20 to 30% you know to refine that tool the way that we want to use it. So we’re taking it very

seriously. Our clients right are demanding that we do and that’s a good thing because it makes us better uh at

delivering value. So we’re we’re we’re there you know we’re just waiting to

see. I mean, if you were to ask me this question, Mark, two years ago, I wouldn’t even not had an answer, you

know, uh, and my answers change about every six 12 months. They get a little bit more refined.

I think that’s that that’s that’s brilliant. Brilliant and honest honest honest feedback there. For me, there was

one thing that came along. So, we’ve been doing AI for about six and a half years now. In fact, believe it or not,

um our our CIO who retired uh last week, he was doing AI in 1994 on condition

part of a PhD. But that’s a long way back. It was AI, it was machine learning, but not the AI we recognize today.

But one of the thing that changed it for me, Monica, which I think is a is a really important evolution is reasoning.

So good AI now has reasoning engines baked in. And those reasoning engines allow the AI to to to tell you how it’s

come up with the conclusion that it has. And that’s very very important. It’s a bit like when you did your mathematics

exam at school or college, if you wrote 42 was the answer, you got no marks. You had to show your workings.

And and and it’s the same principle. How can you trust an AI? I hated that part, too. Me, too. I was never good at it.

Without without the workings. The workings are key. So, so what’s what really unlocked for me as a skeptic, because I’m naturally skeptical towards

this stuff, was when you were able to ask it to tell you how it got there. And I think that that that made a big

difference because suddenly you’re in a place now where your logic, your brain, the way in which you’ve always done it

can now be married up to the logic that the AI was using. And then you can be much more comfortable about accepting or

not accepting the conclusion. So for me that’s that’s been a big thing working with AI providers out there to help

understand how they did that was a real light bulb moment for me. No, I agree. I believe uh you do need to

almost fact check it, which is kind of funny because you would want it to fact check you, but you did touch on an

important point about evolving. So, if you had to give organizations one piece of advice, what would it be around the

risks if they don’t evolve their data? Um that’s one one single piece of advice

if you had you’ll be left behind. You’ll be left behind. Yeah. Eventually, this stuff is going to

is going to be a a a left arm to a right arm. It’s going to be it’s going to work in in in in perfect harmony and and we

have to we have to find its place and adopt it. Don’t be scared. It will it will improve with you. Um it it’s it’s

not a destination. It’s a bit like safety. Safety is not a destination. It’s a it’s a state of mind every day.

Um,000 days LTI free just means you might have an LTI on the 3,000 first day. AI is the same. It’s not a

destination. It’s a it’s it’s a constant evolution. And don’t don’t be left behind. I think for leaders these days,

particularly in large complex operating environments, there is so much that they can learn about their organizations

without putting the human being in uh in harm’s way or without working the human being to the bone. I think it’s a it’s a

it’s a tool that we have to we have to harness and adopt and use use properly. Yeah. I I think the organizations when I

work with clients of intellects, I can tell very quickly the organizations that

are pulling away from their peers in a vertical or across a domain right now

are are treating uh quality and health and safety and environmental data as a

strategic asset, right? Not not some kind of compliance obligation. I would not have said that 10 or 15 years ago,

but they are looking at it very strategically and how they push their process and their programs forward.

Yeah. Yeah. I mean, I think this was a really important conversation to be had. And

just to close us out, um, could you give us both of you, um, a best case example

of what happens when you do everything you guys said? Do you have any success stories to share?

Uh, yeah, there’s there’s lots of use cases and and and success stories. Um

they they happen because of because of a number of things though. They happen because uh the AI is is carefully set up

and configured to deliver the use cases the organization is looking to achieve. The organization is smart about how it

collects data. That’s a really important part of that journey as well. Um and and intellects again has a big big role to

play in that because the way in which reporting data is gathered, the way in which questions are asked, the way in

which natural language is accumulated, the way in which the technical diversity in the data is is is enhanced, that’s

all AI’s friend. So the use cases that are the best ones and we can we can we can c certainly put our hands on many uh

now because it’s been a number of years are always because that that they’ve been carefully considered and carefully

built from the from from the bottom up. The ones that haven’t been successful in our world have been the ones where data

has been thrown over the fence with an expectation that the miracle of AI will just catch it and make sense of it. I

think I think there’s there’s that it’s becoming easier now to ingest more generic data sets without a whole lot of

configuration done in them, but that’s been because a lot of clever work’s been done over the over the years to make

sure that that we’re ingesting and understanding data properly before we turn on the the AI analytics.

industry-wise, uh, aviation, construction, oil and gas, uh, rail, uh,

waste and water, broader utilities, lots of stories where either the the lenses are used to improve safety insights or

potentially environmental or customer service insights or potentially quality insights as well where they’re looking

to reduce non-conformance rates. But yeah, when you get it right and when you follow the steps that you’re are required to to to to build the right the

right starting gate, if you like, then you’ve got every chance you’ll get a very quick return on investment.

And as a practitioner, I I would say this, people are going home safe and healthy every single day.

We are protecting the environment that we are in charge to steward and quality is delivering on its promises. I I think

that that’s it. If we can use technology to do that, we can use data to do that.

I’m one happy practitioner. Well, agreed. That is why we do what we do.

And I say it over and over. Our job is to send everyone home to their loved ones at the end of the day. And data is

your friend. And so is AI if you use it correctly and don’t abuse it to make your grocery list like I do. But I think

the main takeaway for me and people could agree is that even with all this data, the real value only comes when

organizations actually learn from it. So that’s a great segue into the next

episode that we’re going to be having with Comet, which we’re going to be digging into what happens after the

data. So how organizations investigate incidents, we uncover the root causes and what actually turns those insights

into meaningful change. Scott, do you want to send us home? Any final words?

Well, it’s always good and it’s great to be with my friend Mark as we kind of end this and of course with you Monica that I get a chance to work with every single

day. So, thank you so much for attending and we look forward to the next one. Yes, thank you Mark so much. It was such

a great conversation. I’m sure the listeners are going to agree as well. And any final words from you?

No, thank you. Thank you both Scott. It is absolutely always a pleasure to uh to to share a room or a virtual room with

you. your passion and your uh and your commitment to the cause certainly hasn’t diminished despite all those years of

experience. I think I think these things are great. Um you know and and and and

there’s something very powerful about just sitting down having a chat and I think that’s that really fantastic. So

Monica, thank you for organizing and for for chairing a really really enjoyable conversation.

Thank you back and just to call out it has been many years for Scott if you guys weren’t aware.

I’m just kidding. All right, Mark, Scott, thank you both so much and stay tuned for the next

episode. Take care. Thanks, B. Thank you.