Courtesy of Kroger's Dan Whitacre, below is a transcript of his speaking session on 'The Transformation to Intelligent Enterprise' to Build a Thriving Enterprise that took place at BTOES RPA & Intelligent Automation Live Virtual Conference.
You are standing in a room amidst a sea of different unsolved jigsaw puzzle and a pile of loose pieces yet to be connected. Each puzzle must be solved simultaneously while new puzzle pieces are streaming into the room from all angles.
There is no obvious indication of which piece goes with which puzzle, and you only have fractions of a second to decide where a piece may fit into every puzzle in the room and some pieces may fit into multiple puzzles. The time you have to make sense of all this, your window of opportunity, is the time it takes the customer to click a choice in a Digital property, move their eyeball to a new location of a Physical property, or the milliseconds that define a span of time for the next best action.
Now let’s think about puzzles in another different domain, security, are you safe? Who is a potential insider threat? How do you know? These are just different puzzles to solve, the unknowns are the puzzles. How do you answer the unknown and deliver insight that is thought to be impossible? How do you do this with precision and better than your competitors? We will introduce the basics in this discussion.
It's an honor for me to introduce Dan Whitaker and lastly, Brouse, they're gonna talk about the transformation to Intelligent Enterprise. Delaware darker focused on innovation and transformation for at Kroger and its enterprise approaches to data analytics architecture, technology strategies and practices, as well as its disruption roadmap.
He has served as a CTO and Vice President of Business Development for a technologist services organization and spent six years consulting at IBM on eye on information and analytics strategy and architecture.
Wisely roads is the VP of R and D, interest formation for appropriate technology. And digital, he holds academic appointments in cybersecurity, health, informatics supercomputing applications, and advanced R&D research roads has advanced degrees in business and tech knowledge from the University of Houston, the University of Western Carolina, UT, Austin, and ... University of Maryland.
So, Dan, thank you for it for being with us today. We truly look forward to your presentation.
Well, thanks so much, Jodi. I appreciate you inviting us. And I was able to tune in, in some of the presentations yesterday, was It looks like it's a very good webinar so far, And I appreciate the sponsor is helping you put this on, so thanks to everyone.
Today I hope I have to start showing the location.
Today westwards and I want to talk a little bit about the intelligent and enterprising cognitive and contexts computing. But before I do that, since this is, this is an audience from around the world, Both west and I are currently working for copper, you may not know who cover is. So just for a little bit of background on.
brokers probably is the largest traditional grocery store in the US.
On the Fortune 500 scale, we come in at number 40th largest company in the US. We have about $120 billion worth of sales every year, and we serve about over 11 million customers every single day.
So, the intelligent enterprise, cognitive, and contexts computing, or the rise of the sense making machine, we use this term sense, making machine. And, actually, I've seen it in publications much, more often, anymore, is what our ability to acquire data, and to make sense out of it, and to test our assumptions over and over, and over again?
We have to do this in a time that allows us to act relevant.
At this point, I would typically say, play a video, but, as in many presentations, the video works, everything all the way up to the time of the presentation, and now, it doesn't work.
So, let me let me just let me ask you to imagine this, if you will imagine, you're standing on the corner of a very, very busy intersection and cars are whipping by very quickly. And you need to get to the other side of the intersection.
But, the only information that you have is a snapshot or a photo that you took five minutes ago.
How would you cross the street?
Well, this is the way many corporations are working today is, they're looking into the past to decide what to do now.
But as we think of the new competitive reality, the most competitive organizations are the ones that are going to acquire data to learn and as fast as they learn, that is what is going to be very important.
Now, you may ask, why does a grocery store care about that? What why does retailed care about that?
But, if you've been in this industry, you realize we are going through a great deal of transformation ourselves.
Personalization, for every single consumer, endless aisles, changed from consumers wanting things a day from now to seconds from now.
The ability to take the platforms that we have and monetize them, monetize them, meaning I have to change very quickly, to adjust to consumer behaviors.
The need for fresh and local food requires that I keep up to date information on what's available at all times. Moving from five to services, now that I really have to understand how a customer X is much is, what products do they want? So, data science is becoming very important to everything we do, but it's not just data and science, it's fast, data and science.
So, making the new competitive reality. A reality. It's, actually, there's a lot of pieces to this puzzle.
We really don't have time to go through all these pieces this moment. But we're going to touch on a couple important points that we would like to talk about today is managing data as an asset.
Mastering cognitive and context computing, engineering assistance of systems approach and understanding the principles of duality. So I'm going to touch on the data part. Then I'm going to pass the baton to West, who will talk more on the cognitive and the systems of systems engineering portion.
So why does data, man?
Data is the language of business. This is how systems and people communicate with each other.
If the data, if the language of data is not consistent throughout an organization, it causes problems.
Let me give you an example. A single decision up in one of our manufacturing facilities may say I want to ship ketchup.
I don't want to ship catch it in glass bottles. But I want to ship it in plastic bottles.
When someone makes a decision like that the weight of a case of ketchup goes from £30 to 20.
So OK. Why is that important?
Well, when I'm shipping from a manufacturing plant to a distribution center, if the case weight is £30, when we reach the weight capacity of the truck before I ever reach the space or cube capacity of the truck, so I ship the truck half empty because that's all I can put on it.
When the case Mike goes to £20, now I can see much more on a truck.
But if my warehouse and my logistics team doesn't know it now, weighs £20, they'll still load the truck as if it was £30 per case.
Therefore, I'm shipping Air when I don't need to do that.
That problem will go all the way through the supply chain to the point where I want to put product on the shelf in the store. And now the width of a consumer item was two inches. It is now three inches, and I no longer can fit the same number of products on the store.
So, data is a foundation for science, cognitive in context.
The other thing that we've tried to make a point of is Data management.
And the point that data management is not something that is done by application, by application, rather Data Management spans Application, and has to be done holistically. So let me give you an example of that.
I might develop a pricing system and in that pricing system, since I'm a retailer, I'm going to assume everything needs a price. If I'm gonna sell it, needs a price, I'll put that rule into the system or some of them do that.
However, the people who set up the item set up system or develop those programs, know, that's a very complex problem, and they don't have all the information when they set up an item.
So they may not have the price yet, but since the previous system required a price, they decided to implement a rule that says, I'll enter 99%, and that will indicate I need a price.
However, the people downstream at the point of sale system as it exists. So they assume the chicken or the product isn't 99% and they'll sell it for 99%.
So if I'm John $8 shipping for 99%, I saw a check in, but I don't necessarily make money.
So it's important to understand the rules across every application on how we manage data.
The big problem we're trying to solve from a data perspective is this thing we call enterprise amnesia.
I've been in the business for a little bit since the eighties and back in the eighties. It was very easy to manage data because storage was so expensive.
We had to be very careful of everything you store.
We're so careful, we store two digits for a year rather than four digits around for a year or two, can, you know the problem that bombs Over. However, over time, technology change and storage has gotten much cheaper. So now everyone can have their own copy of data, but it's really not a copy updated version of data, because everyone changes it to their use. So what is happening now is data volume redundancy. It's growing fantastically, but the understanding of what data we have, where it's at, and if it's fit for use, is plummeting.
So this huge gap is being created called enterprise amnesia. We have to solve the amnesia problem if we're going to make the most effective use of our beta.
I'm going to pass the baton over here to Wes and let him go through cognitive, cognitive and contextual computing.
Yes, All right, thank you. Thank you so much, Dan.
All right, I'm gonna go fast so we're going to try and make this worth your while by packing in a lot of information.
Now, the Intelligent enterprise is really all about getting a lot of data that matters, figuring out quickly what that date is trying to tell you.
Then quickly detecting the most advantageous thing to do.
A computer can add, We know that.
It can tell us binary black and white answers, but what if we want more?
Is this bottom line of 2.25 plus 2.2, flotte 5, 4 or 5?
Can the computer reason?
And I'm trying to advance this, right?
Again, there we go.
Well, you know it all depends.
Depends on watt.
It depends on things that are not evident here, it's something that's happened earlier, is the context.
The context is yet more data we must understand and correlate to this event, and this event is not this data, it has not been considered yet, so what do we do?
Well this is what our thoughts turn back to Alan Turing in his Turing test you know that Imitation Game movie Can the Interior interrogator tell the difference between an answer from a person and a computer?
When we get into come to the complex we need new ways to simplify the complexity through systems of systems approaches, new frameworks and abstractions.
This is what Alan Turing was doing when he posed the question as to form the discussions on the first principles of the field.
Does a computer think does a computer perceive?
Now, we know there are many disciplines, now there's frameworks and abstractions and advanced technologies, we must consider an integrate to achieve an intelligent enterprise. So there's not time to talk about all of those right now.
So, for today, we'll just assume that a computer thinks, in terms of math, parsis, the current situation within the scope of available current data, the computers perceives concepts through the accumulation and correlation of data over time then associate that data appropriately to the current observations to make highly accurate identification or predictions.
So, let's look at the basics of context accumulation.
Let's think about context as a puzzle piece. So here's a puzzle piece, it has flames on, so what does it mean? Anybody know, I doubt it?
So, let me give you some context. Let's say that this puzzle piece belongs in a fireplace, well, great. It's gonna be a great day, get a snack, relax, and you bought your favorite chair, get a blanket. It's gotta be a great day. Well, what does that puzzle piece was on a house? Now, you have an emergency. Call 911.
But what if that house was on a frame, and it was hanging on your wall? Well, now it's art.
So there you go.
It's something to be admired.
So you can see the definition of context on your screen.
By taking into account all the relevant facts that became a piece of data that you're considering, that's the contacts, all these prior fer facts are the contexts. So let's go a little bit further.
Let's switch gears to cognitive computing.
Machine language is the adaptive learning component of AI, while cognitive is the thinking.
and registration components with orchestration components.
We know from the previous slide that what cognitive thinks on, is the data it's presented with.
If I'm building a cognitive system for situational awareness, that will react to, you know, dynamically changing conditions. I need a closed loop system that will understand and perceive the conditions quickly.
You know, respond and learn. Do it again, and again.
To do that, we need a closed loop system.
We put this all together. We ended up with an AI model that looks like what I've got on the screen here.
The closed loop system that I've used in a diagram is an ulu, Observe, Orient, decide, and act.
That was developed by John Boyd in the 19 fifties, now he was an Air Force Colonel who developed it to describe air combat operations processes.
It explains how to direct one's energies to defeat an adversary in, combat and survive. Now, he's was a pilot.
You know, it's a set of interacting, feedback loops that are kept in continuous operation during combat.
So just think of it as a dogfight. When I worked for IBM as the Deputy CTO for US Federal and NATO, I worked with the Air Force, and then later US Cyber com to adapt this same method, this Cloaked closed loop system to set up the US Cyber Defense framework.
And it worked perfectly.
These principles of thinking about the environment are applicable to logistics, digital, business, or any other competitive, responsive business problem.
To win, you must do these things better than your competitor. You must observe your relevant environment better than your competitor.
Knowledge and power, it all starts with gathering the raw information from which decisions and actions are based on.
You must orient the information to prepare it for decision making.
This is where data context plays a major role if you were accumulated by collating these facts of the past and correlating, associating with the other data elements properly. And they connected them to the situation correctly.
This is also where simulation prediction, unwanted bias, notification, pest, re-experience, risk bias, corporate culture and values are also considered.
Then you must decide and act within a timeframe that matters.
What is the timeframe that matters? That's your window of opportunity?
To win? you must continuously do this faster than your competitors.
You must operate at a faster tempo to generate, you know, to tempo to generate rapidly changing conditions that inhibit your opponent from adapting more quickly than you are.
This cognitive learning model is built for decision clarity, then it's optimized for speed.
The orient step is where many fall down. So, let's take a quick look at situational awareness from a sense and respond point of view.
This is a system of systems approach.
So, this is a kind of a nugget, a small cognitive model, a Model lab, this example, cognitive model, it has certain responsibilities. Some of it will, it will take action on.
Some, it will pass to humans for action.
Some it will pass to other systems, other cognitive systems for action, or it will pass it to them as context to accumulate for other considerations. Input comes from any IOT sensors in a store that counts: Smell, see here, detect water, Determine if a door is open or closed, feel he, or cold, et cetera.
Application program: sensors and trucks news reports, traffic weather, sports outcome, police, and other emergency channels, and so forth.
The cognitive node has several objectives it optimizes depend, dio depending on the situation, the emergency the store traffic patterns. Changing labor and stocking plans, new vendor promotions.
So, it's really optimizing store in, you know, in-store advertising models and so forth.
Its goal is to optimize to a prop or minimize an expense. So there's various scenarios and models used to predict optimal possible outcomes and to generate an appropriately optimized action plan.
You're constantly being revised.
The outcome performance is typically improved with better math, and you know, situational data, an additional context. In a grocery store, many things are going on simultaneously. There's never enough resources to deal with every situation in a timely manner.
Store managers are always making decisions about the next best action based on a wide variety of multimodal signals.
Cognitive systems is about helping them make decisions.
But it's also about making decisions for them because they make it faster than they can. So let's look at this layered together.
Higher functioning cognitive systems are like this, where this is when we did for the virtual store manager. It's a, it also is, when we use for AI, cyber defense systems, you're any highly intelligent learning system.
They're made of layers of cognitive managers, bounded by rules of engagement, rules of engagement, or where you're, you're telling it, this is your, your authorities that I give you to operate within, things You can't do things you can do informed by information that's informed by a higher level AI, they integers, and they operate in this, this continuous set of loops.
Now, the thing that I want you to see through all of this is they all have a different set of functions, and that's the way we break down the complexity. We have sensors and information operational systems that are constantly gathering and looking at the data, and picking up things, making decisions from these the things that you saw in this prior slide, They're all made up of a variety of things that you saw here.
This is just an explode out of a small set of many of these that are just bounded together to do check out a system fraud detection, security, production, planning, and so forth, all coming together and then re orchestrated to achieve to achieve a much higher outcome.
You may be surprised that a grocery store would have so many pieces of information in so many things to do, but that's the complexity of any organization. And we minimize the complex they are humans deal with by getting all this information of relevant data.
And taking care of it for them within within these I AI systems so that the human can really focus on the things that it takes a human to do.
That's really the power of AI.
So let's get a kind of an idea of what a continuous process might look like with the textural referencing added into it.
So this is a sense in response system there inside these loops that I just discussed, you just saw on this prior slide.
So new observation, in this case, let's say it's a person coming out of a store, that's the observation space, is the thing that you're trying to observe.
Now I'm going to bring it back up and show that there was something that we then abstracted from the observation space through a camera, through a sensor, a door, receipt.
Anything with that, that event is the yellow puzzle piece with the little star on it. That's the thing, I'm nubs observation, the puzzle pieces that you see already put together.
The green, red, blue, yellow, brown are all the puzzle pieces that have been identified with this customer before and they're already contextually referenced.
Prior history, prior things.
Now, when we have identified that observation space that belongs to that customer, in this case, we call that date. That's data finding data, that observation, sam, the puzzle, this customer that it belonged to.
Then it has to make a decision.
Do I know enough to, to to accurately put it into context within this contextual store? These are typically no no SQL row stores, by the way.
It makes a decision. You know, I really want to know something else to really snippet in place well so it knows other sensors that it has. So it makes a decision on what sensor it was to activate to get some more information.
Perhaps this wasn't, this was a video thing that it picked out and took an observation, and it wants the receipt, so it asks for the receipt.
And there that with that piece of data, it then was able to snap that, that prior observation in place, that here comes, that that receipt.
And now these relevance, this given directed attention, these are these other folks you see in the directed attention box, they've registered.
Things that they want to know. Tell me when there are pink elephants on the White House lawn? Don't tell me.
No, tell me that until it exists. So they read, they registered things, they wanted to know that when true, let me know.
So, I'm gonna just stay over here, do my thing, and when all these other conditions are true, and this condition happened, this thing occurred. Hurricane in the Gulf, whatever it is, you let me know that these are the things that I find interesting.
Guess what? That blue puzzle piece made something interesting, happened. So they were told, this unique condition you already know about just occurred.
Now, deep reflection if you ever sit in a couch and then you're watching a show and all of a sudden you go, oh my gosh, Harry at work, less, Alex. Just watching a ballgame, and it just popped into my my hearing loss, Sally. Oh, my goodness.
That's deep reflection. You're doing data mining in the back of your mind. You're describing all these facts in the back and all of a sudden some things snapped and snapped into place.
And you realize that those things collectively met.
Harry must love Sally.
Well, that's what we're also doing, These systems, We're doing data mining in the back.
And we bring those data pieces into, context, foursome, or some non obvious relationships. And in this case, that snapped in a new pattern.
That new pattern was then when formed, that group of folks. Those people go, Oh. That's so interesting. I love that, Wow. I'm going to register a brand-new interest that if it becomes true, let me know. So now we've expanded our interests.
So if I look at this, this first element is Extract is normally done in a streaming mode, can be done in batch.
This is empty resolution and relationship discovery. This is machine learning, Data mining components, also deep learning can live here.
And this is grit, graph visualization, case management, and so forth. Scoring, predictive modeling, that processing lead back in this decision side.
And that is that lieut.
Now, I'm going to quickly rotate to identity duality and insider threat. So I'm gonna take these things we just talked about that can run stores and make profit and so forth, and I'm going to use them again.
The Intelligent enterprise, they've got to understand about security.
So let's talk about that real quick in identity duality insider threat security questions.
Quick little definition. You see all four. I'm going to talk about two Whatever it is.
Well, what does dwelling?
It's two things that can be opposite at the same time.
I'm a trusted employee.
I'm also an insider threat, A great suit.
Normally associated with doing business in Japan, but you see that. You see the definition I have on the screen, but it's the thing that I want you to do really snap to its secret relationships.
It's relationship's not out in the open things you don't know reveal themselves and surprise you. You don't like surprise insecurity. You don't like surprises, bad things.
So what I want to talk about a moment and how these these bad things can happen.
So let's go back to 20 14 Guardians of Peace. Y'all remember that story about Sony being hacked?
That wasn't a fun, was it?
That story really cause you some problems, let's, let's think about it. A second.
Hackers not only got in and erased its systems, they got into its computer infrastructure using a virus, some malware, they sto, approximately 100 terabytes of sensitives Sony data. That's a lot of data.
Then they started gradually releasing that data to the public, included, pre-release movies, people's private information, sensitive documents.
They released documents such as employee salaries and bonuses, HR performance reviews, oh yeah. And we're getting sensitive. Criminal background checks. Termination records, correspondence about employees, medical conditions. You know these tactics were aimed at demoralizing Sony's employees and causing them internal instability.
It was to instilling fear and Sony's employees and reducing their productivity.
This was intentional to cause reputational harm and deter celebrities from working with Sony due to fear about possible leads.
If you remember, that was because of the, as it was later, you'll hindsight, it was associated with a country's disdain for a movie that Sony was going to release. All right. So that occurred. Let's go back to the house 2015 Anthem.
Disclose criminal hackers, it broke it into its servers.
So approximately 80 million health records, personally identifying data, OK, not good.
Now, let's think about this a minute, think about the college information that was, has been stolen.
Think about all these other things that have been stolen, those things being correlated.
Now ask yourself this question What if we had all this information from all these leagues correlated together, and you've got this wonderful employee.
And they get a phone call.
They know exactly how to get to your employee, no past behavior of any sort of any bad things happening. This guy was the model of perfection.
He says, I'm going to do this to your kids, I'm gonna do this, your wife, I know you don't care about yourself. If you don't do this one little bitty thing for me.
And they do this to the employees, to this one little bitty.
No small little thing that that person goes, well, that's not that bad.
And then you add them all together, and it's something huge.
But they knew how to get to them.
Suddenly you have.
They just let massive company harn get out.
But each one individually didn't think they were doing much arm at all. So, agency that, was that bad? Did you actually put them together? It was huge!
So, what happens if past behavior is not an indicator of future behavior? They need either zero trust security model and the way that you figure these things out is you've gotta collect up all these little, little, small little hints that are happening in your security system. You need a system response system.
You've got to be able to go figure out all the slightest little bitty hints about what's going on in your system And you gotta change, you gotta change the way you do security. The same system I'm talking about equally applicable here.
Let's talk about duality, all these things all these pictures about duality.
Picture of Dorian gray duality you remember that story? A close duality.
I am a loyal citizen. I am also revolutionary.
Counterfeit goods duality.
I have one, but the other blue and gray. I'm your, your, your buddy, I'm with you, fighting with you, but I'm not gonna get into battle. I kill you, too. You didn't know. I really wasn't with you.
Now I want to go to this, and we're going to go to a poll in a minute.
On Saturday, February the first 1970 fire destroyed most of ..., The ... Companies capability is 23% owned by Toyota Motor Company.
There are a number one factory, which did all of Toyota's break assemblies.
This was bad. Is shut down the plant, shut down to a yoda?
Oh my goodness.
Sorry, I had two suppliers, however 99% of their ..., that were critical for every car they manufactured.
OK, from ace, Mangas impact shut Toyota down, it was terrible.
It was so bad that Toyoda's Production Hall would lead to an unacceptable decrease in Japan's total industrial output.
Now, I want you to.
This is bad, terribly bad. They've got to recover from this.
While we bring up a poll, you vote on what Toyota did after year to mitigate this.
I'm going to read you a little bit more about it. So bring up the Toyota poll.
To avoid this undesirable outcome Toyota and Nissan called upon members of their ..., they just, they just said, We just need to help broadcast it on TV.
Without anything other than this, we need help.
This is what happened, 36 suppliers aided by more than 150 saw subcontractors instantly, separately without communicating with each other.
Fired up, Fired up separate production lines with each production line, uploading small batches of the ....
This was done by family affiliation data to your debts gratitude, non obvious other relationships. Nobody could have mapped it out. Nobody knew what they were.
Brother sewing machine, or sewing machine, was a family member, had a family member, three times removed. They created the veil, although they were terribly inefficient, terribly.
Costs are event. But, Toyota, without any permissions from anybody they paid everyone who helped them, they paid their cost, their lost revenue, they're retooling. They paid $100 million worth of bonuses.
No receipts required, Their word was all that they required.
They finally got up, they got things repaired, but this was a terrible, terrible cost. So, my question to you is, what did Toyota do to make first, to respond to this massive fire, this obvious single?
Oh, reliance on?
This single supplier, So, why they do, where we, at the poll, everybody, figured out what they got.
Government insurance, three alternative suppliers ahead tube at 99%, bounce about, uh, 40% defined it so far.
OK, I'll give you NaN.
3% of Japan's GNP.
Lot of risk.
This is due risk management. I'm gonna give you NaN.
And then I'm gonna give you the answer.
You gotta pick, five, or, three, two, well, let me see the results. What we got?
Yes, you would have thought that they would have gone back to three alternate suppliers.
13 of you said they'd go back to what it was. That's a surprise.
six for government insurance, 21, make sure they had another alternate supplier other than them.
They went back to exactly the way it was because their data showed that they could depend on that caret to come bail them out again. And the cost savings were so much.
It was such a differential activity for them, such a differential benefit that, that was good enough, they would risk it, depend on their ....
Data is power, and that's what their data showed for them.
They could depend on it.
Now I'm going to ask another data question.
I'm trying to advance the slide.
And I want to bring the poll up for the American Revolutionary War, so you bring that poll up if you will.
And I'm going to talk about it.
Because we're all experts we all know what happened in the American Revolutionary War, so I'm gonna ask you this question, consider it.
Which one of these things, if it didn't happen, would it prevented the American Revolutionary War and we have the answer. It is one of those four data tells the story, I want to show you the the impact of data.
We have all the records from the past. We have, we're pounding those records, we've used Network science nodes, erecting in an ecosystem upon each other to go look at this.
And, I want to see how good you are You, You know, you've been through classes, you've been through college, you've been through at least high school in American history And the answers are there. It's written down the history books, you've seen the Specials.
We did a study, we looked at it all, and it and the data was overwhelming. Matter of fact, the History Channel in their In their special the Revolutionary War, made note of this.
I don't want to describe a model I'm going to show you So we're going to take NaN here and let you let you vote Consider it.
I'll show you a model where environment factors. There are factors that are irritates, a three-dimensional model. There are many factors that are factors that irritate you.
They create a feeling of frustration with islands you'll deal with. And you'll accept.
Constraints are factors that constrain your freedom of action in a very material degree, and they change your abilities to provide an acceptable lifestyle.
And in the middle, of all my models, show insider threat activity, And you can always see insider threat, and understand as a window of various power players, and how they're reacting with each other.
So we included land sales, legal activities, every kind of ..., what seemingly, would be insignificant data.
Now, remember, what the Navigation Act was, England's American colonies can only export their data English alejo ships.
They basically just did enriched England.
Uh, remember, pine trees, The pine trees were, that's aneel's ability to build ships and there's only two kinds of pines Google ads of trees.
one of them was to be the mass, and they had to be very, very special trees, and Britain had already cut down all their trees. Spain had the others in the American column. He said the white pine, which was a very specific treated.
Now as we look at the pole, Boston Massacre, yes. Quartering Act and putting people in your house and spying on you, navigation, actually use my ships, 21% of you pick the white pines. So let's go let's go look, Let's go look at the model and get the data.
And then we'll summarize.
So let's let's go past that with the model up.
This is one that was done with doctor Chan. There we go. Go back And you already saw this special, here's the model and the, that has the ways.
We just talked about that latent stability model, and it's, it's the data and the method is good that the State Department was using, using this, too, the method and the thought process to help predict, you know, future Arab Spring events.
Uh, and as I get ready to go to summary, I'll just, bottom line this, because you already saw the other slide.
But, as you get to read and then white, white, lighter color, that is the the intensity of insider threat. that as you rack and stack the data and you look for what the thing is that ignited it.
That if I removed it, we would then still have anger but not revolution. It is, and the drum roll was the white pines act.
Now, they really needed those pine, because they were the only other tree that would fit for those big masks that were on the British ships, and they really needed them very badly for their Navy.
So with the Royal Governor, Sir, John wit worth decided to enforce the, the Pine Tree Law that was done in 90, 722 enforced it.
As you can see, on that chart, is 776, 70, 75 is when he was governor, and he enforced it within that timeframe.
That is what was the straw, the domino that started the other dominoes that if you to pull that out, that would have defused the American revolution.
So, the question is, how much do you think King George? The third would have paid to know this.
The value of data, the value of the system.
So, here's the, here's the first principles, lie, When he takes teamwork, the key point here is it's not a consensus gain when you're doing transformation, and you're doing things that must integrate with each other. You have to have somebody with decision rights.
These days have got to integrate together, it's it's enterprise architecture is a variety of other things. Interact with your data program or AI programs, so forth.
The other thing that I will bring up to you is the third from the bottom.
Is that this, the cognitive complexity conundrum is a system of systems solution.
The, uh, the next second for the bottom is do not split up your data science analytics and BI programs. Use them as pipelines. They have to be co-ordinated together.
If you're going to get us streaming systems, you got to this data is program is vital to you and if you're going to look at the Latin, the last one is the value of a cloud native security program do not do this bespoke. You will fail. It's not going to work for you.
And, uh, I'm just going to give you a half a second to also absorb, the context computing must be mastered for edge compute and prediction.
Uh, is, you have to do it. You're gonna have to get good at it. These things are integral or you will sputter inspect your, whoever got it. Whoever gets these will then beat you in business, if they're your competitor.
So, these bullets are part of mastering the Tools of Disruption, and with that, I'm going to close, and we're gonna go to Q and A And I would like for you to ask Dan very, very, very difficult questions, Pleats?
Now, why, when he said? Very good, very good at that, and then the last, we're going to ask you to stop sharing your presentation. at this point. If you can click on the stop sharing presentation button, then Great to have you. Here. We have, we're just one minute from, from the scheduled time, so we'll be all for the questions here on, as quickly as well, maybe do 1 or 2. Jonathan asked, Dan, if you could flesh out the transformation timeline at Kroger: What digital business transformation timeline look like? A Kroger?
Yeah, I'm not gonna say it's done, because I really do believe that transformation within in a large corporation is a continuous process. Our customers are changing. Our business models are changing cove. It has changed our business model dramatically.
People don't want to go in the grocery store. We have to be make sure that we have a great customer experience. Wow, we protect the safety of our customers and the safety of our associates.
So I don't think there's a big, there's a beginning and end, and quite honestly, if you could just give us some insight on, you know, have you gone through transformation process in your organization? Where maybe five years ago, this started becoming, you know, as a strategic direction that your talk is, this something that you have done the last 12 months.
Just just some rough timeline on when you started down this path that you share with us, it's probably in a couple of years, then some of the ideation and thought processes and so forth, and I agree with that.
Very good, another question, Cameron was asking about a set, that was a great example of cognitive cognitive system for a grocery store, and he just curious about, if you take into consideration pricing and promotions as a factor on the, on the on the model for the grocery store that you showed, the data.
Let me read that one.
That is just one of a ton of other considerations, We did not put in the, in the model, because we did not want to show all that. But, just think of every consideration that is important to a customer, or optimizing a customer experience, or a differentiator that we wish to add to the experience. You can continuously change or remove those onto. That is just a another no consideration you add, long as you have the data. That's three exams.
Very good, very good. And 1 last 1 here, from, from Jonathan, on the virtual store manager. He's asking, can this cognitive tool be automated to run as a single bot, or is it already?
I'll answer that one.
You could, that wouldn't be advisable because then it's difficult and complex to change.
So, so, here's just kind of, one of the resilience factors you want to do is put architectural boundaries and capabilities, So when you get into, uh, the kind of programming to where you're changing and automating mini mini mini mini capabilities, the way you reduce complexity is make functions self contained.
So now I can go in and update that function without impacting many other functions. So that allows me to do, like, you would see on your phone, where you can go and have all these different changes to the phone, and see them updating daily, or once a week. And you do so instead of having this every six month release. If you put them all together. You're going to, you're going to have this one humongous release that has got a lot of risks to it. So that's also a risk management methods to make them highly specialized and highly componentized.
Farewell. Wes and Dan, thank you so much for taking the time and sharing your experience and the wonderful journey. The ... has been on with all of us today. We really appreciate that.
Thanks to you. Thanks, everyone, and your audience and your audience, do appreciate the opportunity.
Thank you. Thank you. Ladies and gentlemen. This will finish this portion of our session. Couple of reminders for you. As you close out of the session, close the box for the webinar, and there will be a close button that, if you click on, there'll be a short survey. And any feedback that you have regarding the session, you can answer it there. We have on June 23rd through June 25th, we have I BPM Live, that will be the next event for The Business Transformation, Operational, Excellence Live Summits, so that, you can register for that at I BPM live dot online.
IBPM live dot online, and it will have a terrific sessions with amazing hosts, amazing speakers, on very practical applications for intelligent BPM specifically. So, we're going to be finishing up this session. I will see you at the top of the hour, five minutes, to the top of the hour. We'll restart the session. And that will have Margaret Shimomura, Director of Quality and Transformation, and Morningstar, talking to us about controlling robots, how you design an effective RPA governance program for a large financial services organization. So, you do not want to miss that. I will see you in a little bit. Thank you.
Senior Director Transformation and R&D,
Senior Director Transformation team charged with engagements to enable the transformation of customer and associate experiences through the harmonization of strategies across business units, technology departments and external partnerships and a value-based approach to utilizing data, science, technology and engineering methods. Asked to re-join Kroger in 2015 to lead the initiative that would transform the organization to an enterprise approach to data and analytics architecture, technology, practices and strategy.
In between Kroger tenures, Dan served as the CTO and Vice President of Business Development for a technology services organization transforming to a data and analytics consultancy eventually acquired by Data Intensity.
Additionally, served six years consulting at IBM on Information & Analytics strategy and architecture in multiple industries including retail, healthcare, manufacturing and U.S. intelligence. While at IBM, Dan served in the IBM Federal CTO office, the Information and Analytics Tiger Team, the Big Data Champions team, as a client software advocate for IBM R&D, the IT Architecture Certification board and the Information On Demand Executive Architecture team.
Search for anything
October 28, 20211
1:00 PM - 2:00 PM ET
November 9, 2021
11:00 AM - 12:00 PM ET
January 13, 2022
1:00 PM - 2:00 PM ET
January 27, 2022
1:00 PM - 2:00 PM ET
Watch On-Demand Recording - Access all sessions from progressive thought leaders free of charge from our industry leading virtual conferences.Watch On-Demand Recordings For Free
Courtesy of DC Government's Ernest Chrappah, below is a transcript of his speaking session on 'Going Digital To Enhance The Customer Experience' to ...
Courtesy of 's Anu Senan, below is a transcript of his speaking session on '' to Build a Thriving Enterprise that took place at Enterprise ...
Courtesy of Tasktop's Dr. Mik Kersten, below is a transcript of his speaking session on 'Project to Product: Driving Digital Transformation Insights ...
Courtesy of Nintex Pty's Paul Hsu, below is a transcript of his speaking session on 'Improve employee productivity during and post-COVID by ...