FAQ Unpacked. Answering Your Questions with Ryan Sullivan

This is Making Better Decisions. I’m your host, Ryan Sullivan. Decisions are where rubber meets the road for organizations. Each week, we’ll be learning from people who are on the front lines of turning raw data into better outcomes for their organizations. This show is sponsored by Canopy Analytic, helping companies make better decisions using data.

Ryan: Hey everybody. Welcome to another episode of the Making Better Decisions podcast. Today’s guest is me. So today we’re actually doing a little bit of a shakeup. Today we’re going to be answering some of our most frequently asked listener questions. Since it’s just me, I want to take an extra second to thank all of the listeners for making the podcast a success so far.

I generally don’t like being on camera or on a microphone, but This has actually been a ton of fun and I have gotten to have some of the best conversations with a lot [00:01:00] of really cool people. And it is the listeners like you that make that a possibility. So thank you guys so much for listening. Uh, to that end.

Kind of want to keep the train going if we can. Um, if you like the podcast and you’re finding it valuable, please make sure to tell a friend, uh, leave a review, um, you know, like, or subscribe, you know, all of that kind of normal stuff that people that make content ask you to do. It really does help to keep this going.

So without further ado, let’s get into the questions. Alright, so the first thing that I thought made the most sense is I should answer the same question that we ask everybody else, which is, what is one thing you wish more people knew about using data to make better decisions? Tying your data work to a clear business goal. This is something that we’ve heard on the podcast quite a bit, both for me and from a lot of the other people that we’re talking to, [00:02:00] but organizations have a purpose. If you’re a nonprofit, it’s to achieve some sort of, you know, mission. If you’re a for profit company, it is to make money.

And Typically, we have specific ways that we’re going to go about doing that. So, you know, if I’m a, uh, hardware manufacturer, like I make nuts and bolts, right? Like going out and like making a phone app, you know, like, obviously there’s, there’s specific things that we do. So we can set up goals along those lines, right?

Well, okay. We want to make, uh, The highest quality screws or we want to make the cheapest screws or whatever the case may be. You can set out those goals in line with your organization’s purpose. Data and all of the tools that we use to interact with data are just things in our tool belt to help us achieve those business goals.

So if You know, building a phenomenal, awesome data warehouse or building some [00:03:00] sort of new AI model or all these reports. Like if we can’t very clearly say how that’s going to help the business achieve those goals, maybe it’s not the right project to invest in. On the flip side, if I know that I’m trying to go about achieving that goal and I’m like, Hmm, I, I just, I’m at a sticking point.

I don’t know how to make this decision or we have to make decisions like this very frequently and we always feel unprepared. That’s a great sign that I’m on the right track. If I know exactly how the work that I’m doing is going to improve the business’s functioning or its outcomes and decision making, that’s how we know that this is a good thing to invest in.

All right. Thanks. Now, the next question that we had gotten in is how do you balance the need for thorough data analysis with the need to meet tight deadline? This is a good one. So obviously in business, we’re always, you know, optimizing for scarce resources. And [00:04:00] so you could fill up any amount of time.

with with making data analysis better and more comprehensive and all this stuff. What I really think that it comes down to is what I talked about in the last question, like what is the actual goal? What is the decision that needs to get made? And then we can have a discussion with, you know, the person kind of making or doing that analysis and the person that, that needs the answers.

It’s like, you know, okay, well, if the decision has to be made on Friday and it would take me until next Tuesday, To do an analysis that’s helpful. It’s like, well, maybe, you know, we gotta, we gotta change some things here, but you know, usually there, there are parameters around this and you basically just want to want to have a conversation and communicate with everybody, Hey, here’s what we have.

Here’s what we don’t have. Like these investments have or haven’t been made that prepare us to be able to do the analysis. And then you go from there. So just communicating across the [00:05:00] team about what are the priorities? What are the, the, the, kind of realistic timelines to do the different pieces of the puzzle.

And then you kind of come up as a team to make a decision on, you know, what is the depth of the analysis that you want to do? You know, how much data do you want to pull in? Like anything along those lines. All

right. What is the best way to start building data culture? So there’ve been a couple podcasts guests that I’ve, I’ve gotten to ask this question to, and My personal take that was certainly informed by some of the things that they said is, is that culture is all about people. I am a firm believer If you have a group of people and you don’t change that group of people at all, the culture can’t change.

Now I can add a person and change the culture. I can subtract a person and change the culture, but it’s really hard to say, Hey, I’m going to take these [00:06:00] 10 people and like materially change the culture of those 10 people. It’s really kind of the sum of the people in the room. So for me, when we’re talking about building a data culture, What we’re really talking about is building a collection of people.

there are really two ways that you can go about this, right? Like maybe you’re building a team, in which case you’re going out and you’re finding people. I think the biggest thing that I would focus on if you’re going out and building a team is having Diversity of thought, having

people that are collaborative and, you know, in, in really any field, you can have these like super duper high performers, but they’re the kind of lone wolf type. Um, and in general data is always like very much a community effort because you have the people who are generating the data, the people who are processing the data, who are, you know, analyzing the data and the people who are [00:07:00] consuming that analysis.

Like lone wolf doesn’t really play great in that culture. You got to work really well with other people. Um, and then the last is kind of this like blend of technical competence, lifelong learner types that are always picking up something new. If you go out and you kind of hire those types of people, I think you’re probably going to end up.

In, uh, in a good spot. Actually, one other thing that I should mention on that is communication. So being, and of course everybody, you know, it’s always about communication, but I think that. The very specific type of communication that I’m talking about is having the ability to explain technical topics non technically.

This kind of goes back to that kind of collaborative team nature to data. You are always going to be interacting with people that don’t know as much about data or technology or analysis or programming or whatever the case may be. As a person on that team, but they’re still critical to engage with. I can’t just ignore those people.

So [00:08:00] having that, that communication makes for, you know, much better collaboration. Now, the. Equally important thing to talk about when we’re talking about building data culture is let’s say I already have a team and maybe we want to do more data analysis. We’re going to start investing in this more. Maybe we recognize that our data culture isn’t ideal and we want to try to do some things to improve it.

It’s still all about the people. Now there’s like a certain handful of roles that are critical to any data function. Right. So you have whoever it is, that’s kind of like handling all of your, whether you want to call it, you know, uh, data engineering or ETL or in, uh, integration or whatever, right? You have the people who are taking data from all the different places and moving it around and transforming it and getting it ready.

You have the people who are doing the analysis. Now these people might be in. An analytics department, a BI department, uh, FP& [00:09:00] A, like any, you know, you can put them anywhere, but these, these are the actual people who are going in and like digging around, building reports, understanding the business, blending that with understanding the data.

These types of people and a handful of others, right? Like people that do DB dev work or, you know, IT support or any of those things, like these people. Just are a part of the team, right? You can’t really have a data culture that excludes people that do core data work at your company. But the thing that’s really exciting is going out and finding the data diamonds in the rough.

So what I mean is that sprinkled throughout this company are people who either know or don’t know, but have the capability to become champions for database decision making in their area of the company. So this might be a salesperson or sales manager, somebody within the sales function that really, really likes digging into the data.

And you can work with that person and kind of make them a champion. And then that’s a [00:10:00] bridge in between some sort of centralized analytics function and the sales organization. And the same thing goes for any of the other departments, whether that’s, you know, finance or supply chain or whatever these other parts of the company are, the really, really fun part is like going out and finding, you know, if you’re a data person and kind of finding your people hidden throughout the rest of the organization.

And what I think is really important is not thinking like, Hey, I need to kind Convert people, or I need to like bring this person into the centralized function. It’s like building all these, you know, spokes of the wheel going out into the organization and then empowering those people to do the right amount of stuff.

So maybe I find one person in finance who is phenomenal and I can give them a ton of access. I can allow them to do a lot of self service. And then the flip side is I might find somebody in another department who’s like really excited about stuff, like wants to learn. But maybe I can’t give them as much raw data or I can’t, you know, I’m going to need to do a little bit more of the work for them.

So like, figuring out, okay, here are all the people, [00:11:00] here’s, there’s like, here are their levels of capability and interest and growth potential and all of those things. And then, Giving those people the opportunity to grow, to develop those skills when they want to, giving them a pathway where they can learn more about this stuff, where they can interact more with, you know, IT or data or whoever, you know, is, is kind of running the centralized data function if there is one.

Um, and, and for me, that’s what I’ve really seen data culture flourish is when you’re finding these people and giving them a pathway to learn more and to contribute more. That’s what gets people really excited. Okay. So we have a couple that are pretty similar to one another. I’m just going to read all the questions and then kind of treat them as one.

So what is a data model and why is it valuable to the business? Where should a business start with data engineering? And then when do you switch from Excel to modern reporting platforms and databases and things like that? So, I want to [00:12:00] handle this a little bit as a, as a story. I think that there’s a, a natural grow up story, kind of like an analytics maturity path.

Um, I don’t want to make it sound like there’s just one path cause obviously there are all sorts of different changes that I would make to this for an individual person or an organization. But in general, I think that most organizations start with, uh, like a single question needs to get answered. So either the person that asked it or they find somebody else, they go and they, they figure out the answer to that question.

So that typically involves, you know, you go into some system like a CRM system or an ERP or whatever, um, you know, the backend of a website, whatever. And, you know, Sometimes it’s more than one and they pull the information that they need from one or both of these and then they kind of do all the cleaning and slice and dice and they put it together and they come up with an answer.

Bada bing, [00:13:00] bada boom. Then what happens is somebody’s like, hey, this was great. So now that’s the monthly report. We want to look at this every month. But it was kind of manually done. And it required a lot of cleaning and stuff like that. So when that starts to become the case or the norm, you start looking at some of these reports and you can start saying, hmm, okay, well, if I’m looking at the same thing every time, should I be doing the same manual steps every time?

And that’s when we, when we can start looking at automation. So whether it’s, um, you know, automating, some of those transformations You know, pull the data and get it ready for analysis and merge it together. Um, or whether it’s, you know, structuring how the analysis looks or any, any of those things like automating that starts providing a huge amount of value.

When I was initially an analyst, I spent, I would say, Probably shamefully around like 90 percent of my time preparing information for analysis. And then I got that like fun Friday [00:14:00] afternoon where I actually got to like dig in and see cool new things about the business that could actually like change and impact things.

So I think it’s really flipping that narrative on its head, where if we can make an investment in some of these impactful to the business, get used a lot, and we imbue them with all of the business logic that they need so that. You know, our analytics professionals can actually be doing analytics instead of data preparation and cleaning.

Then, you know, hopefully we’re getting a lot more insights. So that to me is what kind of pushes this grow up story is we’re looking at efficiency. We’re looking at, Hey, something’s already been done, but we want to make it improved. We know that it’s valuable to the business. So I personally am not a big believer that.

Hey, we’re going to start a data initiative. Let’s do all of this money on data engineering. We’re going to build this amazing data warehouse project. Don’t get me wrong. I have quite a few data warehouses and some of them have been [00:15:00] company changing. However, the pathway to get there should be one of incremental growth in the direction of something that provides return on investment for the business.

So if we have somebody smart, that’s doing a lot of manual repetitive work. Well, let’s automate that work so that that smart person can generate more insights. If we have 15 different data sources that describe our business, and we’d like to actually have an end to end picture of how our company is performing.

Well, okay. At that point, maybe centralizing it makes a little bit of sense. So whether we do that in, you know, one big, Power BI reporter. We build a database or a data lake or whatever level of technology we’re using. I think that every single step should be like an incremental one that shows return on investment and that’s helping us to meet our company goals because otherwise we can really have runaway spending and you know, it, it, it erodes trust in the data function because people wait, you know, you’ve been spending how much for how [00:16:00] long?

I’m just like, Moving this data around, like what are we able to do different or new now? Um, so I think that that kind of natural grow up story is is much preferable. Ooh, this is a this is a feather ruffler Um, what tools and technologies do you prefer for data analysis and visualization and why? I have worked with Microsoft technology for most of my career.

Uh, the main reason for that is the first company that I worked at. That’s what they used and that’s what I got good at. And I thought it was really cool stuff. Um, I do think that around that time, you know, Microsoft and PowerPivot and PowerBI had been building some, you know, Frankly, like really, really incredible tech.

I think that I won’t get into the stuff that’s that’s super under the hood, but the under the hood technology that exists in those tools is really, really phenomenal. I, I think I made a, a huge and, uh, accurate bet that that was going [00:17:00] to be pretty revolutionary to the market. However, there are a lot of good tools out there.

that either compete or are adjacent to some of the main Microsoft ones that I use. So for one, I think, you know, having

the different layers of whatever it is that you’re trying to do nailed down matters a lot. And then once you know, okay, well, we need each of these things to get done. You also need to think about like what are the business’s priorities? So for example, You know, if you’re a shop that uses basically one technology or one technology provider, and that technology provider has an add in that’s going to solve your problem, I don’t really think it makes a ton of sense to go with something else, right?

Like, being a Microsoft person and being familiar with that, right? Like, let’s say that you use, you know, Dynamics and SQL Server and all of that stuff, coming in and saying, [00:18:00] Hey, well, I think you should You know, a different analytics tool on top of that. Well, it doesn’t really make a lot of sense because Microsoft’s made it really easy for all of their stuff to work together.

So I think that there are definitely different things for different organizations that can push them one way or the other. Um, and you, you want to look at the big picture with that stuff. So. Looking at a lot of the different layers, I think the first one is, okay, so you have your source systems. Many times, you know, data people aren’t in any control of that, right?

Like other parts of the business just go out and they procure what they want, whatever they think will help them meet their objectives. And they, they do it. So you’ve got these source systems, whether we’re going to go in and we’re going to like directly pull information out of those into some sort of reporting layer, whether we’re going to try to, you know, you know, Pull it all into a database or whether we’re going to like it and then do some transformations, like however far up the, you know, maturity curve, you want to take it.

The first step is always [00:19:00] just going and grabbing the information that you need. So if you’re not familiar with it, this is a process called ETL extraction transformation and loading. Sometimes people do the loading first and call it ELT. I think that acronyms in general just make things confusing, but, um, that’s what it’s called.

So sometimes, you know, This is also called integration. As I mentioned before, data engineering, lots of big, fancy terms to make, uh, make sure that people think we’re smart. So, which I’m sure we are. So one of the things that you is important about that is the ability to connect to lots of different data sources, hopefully easily, and the ability to transform that information.

So I think that there are, depending on what you’re connecting to and what types of transformations you’re doing, there are a couple of good options. So if I’m just pulling something into Power BI, well, Power BI has Power Query built into it already. If I’m going to be dumping something [00:20:00] into a database or into a data lake.

Then maybe something like Azure Data Factory, which is kind of the, the modern SSIS, if you’re familiar with it. Um, or maybe I want to do a lot of like really funky transformations in, in which case like something like a Python running on, uh, Databricks might make more sense. So we really want to think about the level of complexity.

What are the organizational parameters? And then, um, You know, what are the other pieces of the puzzle that, that, that has to fit into for, for databases? It, it depends a lot on size. My personal take is like, you know, like, uh, there are cool open source options. There are cool options from every major technology provider.

Um, there’s like massively parallel processing options. Um, so I really think that. In this regard, like pick what you’re familiar with, pick what fits in with the rest of your stack. So, but the, [00:21:00] you know, there are a lot of good technologies that are out there for kind of database and storage for visualization itself.

I think I’m going to be the most biased here. I think Power BI has just kind of run away with the front runner spot as far as visualizations. There are other good tools out there that people can use to make better decisions. And that’s much more important than what tool you do it with. But I think Microsoft has really seen this as an opportunity and has invested a lot of money.

Um, and that investment has been successful. And what they’ve done is they’ve really generated a platform of tools that allow, uh, for data analysis and visualization and sharing and governance, all of the pieces that go into, you know, this kind of layer. Um, I see them as a, as a really, really strong front runner, but I’m, I’m obviously a little bit biased.

Okay. So on to the next question. are the first steps an organization should take to build a robust data strategy?

[00:22:00] I’m not entirely sure that they should. Maybe that’s a hot take. Like we started at the top of the episode, hopefully whatever organization you’re a part of has a strategy. Like you have a vision, you have some goals, you know where you’re headed, you hopefully have like a high level roadmap of, of, of how to get there.

That’s the strategy, right? Like I think the, if I take this question, I like start like replacing some of the words, right? Like, you know, what are the first steps an organization should take to build a computer policy? It’s like, We use computers where we’re planning on using them. Computers are great. The internet, man, I’m telling you, it’s going to be huge.

You know, I think that data is just a tool that we use to achieve our business strategy. So Now that I’ve like made that distinction, [00:23:00] I do think that I still understand what some people are talking about when they talk about a data strategy. It’s like, okay, well we, if we make the assumption, and this is a big assumption, but if we make the assumption that we have all the pieces of our company swimming in the same direction to try and achieve a corporate strategy, what role does data play?

in that. And if we call that a data strategy, I’m like, cool. Yeah. I’m definitely on board with that. That definitely makes sense. However, that also provides us, I would argue like a massive amount of direction for what that data strategy is. So if we already know that this strategy of the business is to make screws that are the most lightweight, high precision screws that a company can make.

Okay, awesome. Well, how can We craft a data strategy. Well, it’s basically about answering all the questions in the business [00:24:00] strategy. if we’re going to try to answer all the questions of that business strategy, some of the questions that I might want to answer are, okay, how are we going to accurately measure the weight of the screws?

How are we going to do quality assurance to ensure that they are, you know, made to the tightest tolerances? How are we going to address the, you know, The waste that comes from that. How are we going to look at our supply chain to ensure that we have the right materials? How are we going to look at our, uh, headcount and ensure that we’re training the right people and

that they have safety compliance.

And we have all of these different, what are all of the pieces of the puzzle that go into achieving that goal? And then we say, okay, well, if we know, if we’re answering all of these questions, That gives us all the knobs that we need from a data perspective to make sure that our business goal gets hit.

Then you just set out and say, great, now, tactically, how do we build each of those? Well, we don’t know if all of our people are compliant with safety. Alright, well, let’s plumb into that system or let’s start gathering that data. You know, if we, you know, uh, maybe [00:25:00] our quality assurance is just done manually.

It’s like, okay, well, how can we possibly improve that process and start capturing some data from it and then do some reporting on it? So the, the strategy to me for data is dictated on like, what are the numbers that I need to know to keep my finger on the pulse of the business strategy? And then you just back into, okay, how do I make that happen?

That’s the data strategy.

Okay. What role does executive leadership play in successful data initiatives? I have seen projects where there’s not a huge amount of, you know, very high executive level leadership involvement. however, I have never seen one happen successfully where there is not at least support for the idea.

that leader, whoever is the person that’s kind of championing the project, signing [00:26:00] off on it, if you will, that person has to believe that this initiative can actually improve outcomes for the company. That is the bare minimum table stake. , the executive leadership team focuses on making the project happen just the same as anything else. You know, if there’s something that’s kind of a back burner priority and nobody really has time for it and things like that, like It, you know, it just doesn’t move as fast.

Now it may be that a particular data project is not the highest priority and then it shouldn’t be treated like it is. But if what we want to have is a successful data initiative that meaningfully impacts the business in some way that the business actually cares about, then that should be a priority.

And if it is a priority, let’s clear the way for that. Let’s make sure that everybody. That we need is in the room and that it’s properly prioritized for all of them. And then if there are objectives or roadblocks or problems, the [00:27:00] executives use the power that they have to try and clear those and move the project forward.

That’s what makes things move the fastest and get completed. Um, with the highest efficiency, the most success.

So here’s an, an interesting little series that we came up with. So we had some questions that were specific to, uh, specific, uh, industries within business. And so we thought that it would be cool to kind of like put all of those together and try and rapid fire some of those. So. The first one that we have is for retail.

So, how can retailers leverage data to optimize inventory management and reduce waste? That’s a great question. I have done a fair chunk of work in retail. Um, and, um, I’m going to actually expand this a little bit and I’m just gonna, you know, not just inventory management, but let’s assume that we’re [00:28:00] managing the inventory along multiple stops.

So let’s expand this to talk about maybe the entire supply chain. I think that the biggest problem that exists in this space is kind of the magnitude of connections that we need to look at. So for, unless we have the most simple business, we’re usually keeping track of many different pieces of inventory, sometimes in many different locations.

And The inflows and the outflows from like, if we imagine like a bucket or a pile, the way that lots of people talk about inventory, you know, shelf, whatever, we imagine the, the inflows of new materials coming into that inventory pile. And we imagine outflows of hopefully, Great and very profitable sales coming out of that inventory pile.

It starts getting very complicated simply just through numbers. So the, the amount of things that I have to manage. So [00:29:00] one of the things that I would suggest when we’re thinking about supply chain, inventory management, retail at large is to come up with a couple of pieces that work for all of the piles that you’re looking at.

So for example, how can I describe Predictably for all my products in all of my distribution centers or, you know, sales, uh, locations, how can I describe what do the inflows look like? What do the outflows look like? Do I have a certain safety stock number? And then based on that, you can separate out a lot of the signal from the noise.

So, you know, If I have a safety stock of a hundred units of a particular product and I’m predicting my outflows over the next period to be 10 and I’m predicting the inflows to be 10 as well, I kind of just don’t need to see that. It’s kind of status quo. The types of things that I want to look at Are when I’m shipping 500 units to a [00:30:00] location and the sales forecast only shows five or vice versa.

What if we have massive orders or, you know, we’re talking about umbrellas before a rainstorm or, you know, some sort of. Idea where we can look at an imbalance between the inflows and the outflows relative to the amount of safety stock that we want to have at a given location. Um, and the better that we’re able to do that, the less waste that we have of like shipping things around to multiple places or, you know, spoilage, if it’s not something that can sit on a, uh, on a shelf for a long time, but I’d really say break it down into those three components and do so in a way that it works for all of your products and all of your locations, and then you can just apply that and filter out.

Only the situations that you need to look at to act to improve. Okay, next up, manufacturing. How can data analytics enhance quality control and predictive maintenance in manufacturing? Um, I’m going to handle predictive maintenance first because it’s much, um, More complicated. So I’m going to give it a shorter treatment, actually.

So [00:31:00] predictive maintenance is, I would argue, one of the coolest applications of machine learning models that we had seen prior to generative AI. So being able to predict ahead of time, like, Hey, this elevator we think is going to break in the next month. So. You know, let’s just automatically call up the elevator technician and they come out and they fix the elevator, uh, even though it’s not broken, but now it’s not broken and kind of reducing downtime with, you know, in manufacturing construction, a lot of fields, like what we’re really optimizing for is, is, uh, a minimal amount of downtime.

So if we can do that with predictive maintenance, that is incredible. You can build models that will look at that at any number of levels of complexity. So it could be as simple as, you know, we do predictive maintenance with our cars and saying, okay, well, every X miles, I’m going to go get an oil change.

Well, depending on the temperatures that you drive in and how you drive your car and all that stuff, like, you know, 3, 000 miles or whatever. That might be a good estimate. It might be short, might be [00:32:00] long. We don’t really know, but we’re just kind of using that to be predictive about like, Hey, I don’t want my engine block to seize up.

So I’m just going to do it at a time where I know that it won’t. You can do that all the way up to building like unbelievably complicated machine learning models that incorporate tons and tons of different factors. So it. Pretty much just depends like how important is this to your business and how much do you want to invest in it to incorporate all the information that you can, uh, for manufacturing.

This makes a lot of sense for, you know, some other stuff. Like I’m just going to change my, my oil every 3000 miles. That’s pretty simple. Um, no, as far as quality control. I think the data analytics can obviously enhance quality control because you can find, if you can tag the different types of defects that you’re seeing in the quality control, then you can start doing root cause analysis.

Okay, well, which pieces of our assembly process or manufacturing process touch the piece that had the defect. What [00:33:00] type of defect is it? And then you can start doing root cause analysis again, you know, big in manufacturing is reducing downtime and reducing waste. So if you can do that root cause analysis, it will, you know, massively improve the profitability of that one or floor.

Okay. So telecommunications, what are the key benefits of using data analytics and telecommunications? Um, so, I worked in telecom for my first job out of college. And the biggest thing that I’ll say there is there are a lot of phone lines and those phone lines make a lot of messages and a lot of phone calls.

So data quantity, um, was really the biggest thing there. So getting Um, tools that can handle that volume of data was really what was critical. So when we could actually see, Hey, very predictably, what are we buying? What are we selling? What are, how are people using it? It allowed us to ensure that the. You know, there was less breakage that [00:34:00] we were purchasing the right products, that we were selling the right products to our clients.

Um, it was really, you know, just a massive amount of information to sort. So it’s really like looking at some of those modern technologies. That’s what actually pushed me into learning that PowerPivot was hidden in the back of Excel. And that kind of started the whole journey that I’m on, um, learning about databases and, you know, things that were much larger than the Excel files, because I just couldn’t crunch all the data.

in there. All right, so it looks like we have a little bit of time for

energy. How can data analytics help energy companies optimize their operations and reduce cost?

I’m going to just narrow my focus to a couple of good use cases. This is certainly not everything. Um, and I also, you know, while I’ve done a lot of work for public utilities companies in the oil and gas sector and all that stuff, I’m by no means, um, you [00:35:00] know, an expert in energy or for that matter, really anything other than, you know, the, the tech that I do consulting in.

But, um, one of the biggest things that I found in the energy. sector is same as manufacturing construction, like downtime of equipment and processes is critical to manage. And the other big thing for energy is how good of a picture can I make of what demand is going to look like? The big reason for that is When we talk about energy levels at kind of the grid level of magnitude, we’re talking about states or regions or countries.

There just isn’t effective data, uh, uh, energy storage capabilities that could meet it, right? So I can’t generate You know, if we imagine like grid hours worth of energy, [00:36:00] right? I can’t generate 10 grid hours worth of energy at a time that’s convenient or cheap to me, and then store that for six months and use it later.

I can store it for a little bit of time, And I can balance things and I can like buy and sell to different grids that I might be connected to. But ideally, what I really want to do is I want to have a pretty good picture of when demand will spike and if I can, how high it will spike. And that will allow me to line up the production to meet that demand so that we don’t have outages, so that we don’t have, you know, overuse or so that I don’t like produce a bunch of energy that then, you know, I, I can’t.

Use with anybody on the grid and I have to sell for pennies on the dollar off my grid. That’s, those are really the two big areas, like all of your equipment, right? Anything in energy, whether it’s like I do offshore oil drilling. Okay. Well, you have a ton of equipment that you need to make sure is constantly drilling oil.

If you have [00:37:00] a renewable wind farm, every time one of those wind turbines goes down, it’s not generating any energy. So like the managing downtime and then seeing, you know, obviously no one can predict the future. Humans are really bad at it. And data and AI, I’m not sure. Uh, I guess the, the jury’s still out, but.

can really only do a limited job of predicting the future. But if we can figure out things like, Hey, well, every time it gets hot, people use more air conditioning, air conditioning uses a lot of energy. So if we know that the temperature is going to go up in two days, maybe let’s get everything that makes energy cranking, you know, just things like that.

We’re trying to predict the demand to the, to the ability that we have. Um, okay. So, I would be remiss if I didn’t wrap this up by, um, kind of giving the same little background on me that I like to get from other people, um, when they come onto the podcast. So what do I like to do outside of work? Um, I love riding my bike.

Um, so I’m very involved in the [00:38:00] Pan Mass Challenge, uh, which is a fundraiser that fights cancer. In fact, this year, we will have passed cumulatively 1 billion raised to fight cancer. All Rider raised money goes directly to Dana Farber Cancer Institute. And it’s, uh, it’s really one of the, I stumbled onto it just cause it was a really long bike ride and it’s, it’s really become a part of my life.

It’s something that I’m really passionate about. So please check out the Pan Mass Challenge or reach out to me if you’re interested in donating or riding or volunteering. I also love playing golf. It doesn’t mean I’m good at it. In fact, I’m pretty bad, but I do enjoy it. Um, and then I also, um, I love things that are aviation related.

So I’ve been a skydiver for seven or eight

Years. I have over 500 skydives. I, you know, yes, I fly the little squirrel suits. No, I don’t fly them close to the ground and through little arches. Um, I also am a student pilot. I am hoping to finish up my, uh, pilot’s license in the next year or so. Um, but I just, you know, love.

I think that’s a lot [00:39:00] of fun. And when I’m not doing that stuff, I love hanging out with my wife and dog. Um, the best way to reach out to me so you can check out. Um, me on LinkedIn, uh, just Ryan Sullivan. Uh, you can also look at my web, my company website, which is canopy analytic. com, uh, who are also, you know, as, as you heard the sponsors of this podcast, um, and with that, I’d love to close up.

So there’s no point in me. Thanking myself. Um, but I do want to thank all of you, especially if you listened this long and you got to the end. Thank you guys so much for listening. Thank you for making this podcast a possibility. If you learned something today or you laughed or you got anything out of this, please tell somebody else about the podcast.

Please make sure to go on and give us a review or a follow. Uh, anyways, so without further ado, this has been another exciting episode of the making better decisions podcast. Thanks for listening.

Outro: That’s a wrap for today’s episode of making better decisions for show notes and more visit, [00:40:00] making better decisions dot live a special thank you to our sponsor canopy analytic canopy. Analytic is a boutique consultancy focused on business intelligence and data engineering. They help companies make better decisions using data for more information, visit canopy analytic.

com. There’s a better way. Let’s find it together and make better decisions. Thank you so much for listening. We’ll catch you next week.