What is Business Intelligence and what should I know about it?

Oracle Academy Tech Chat

In this episode, I’m joined by Cathye Pendley an Oracle Ace and Business Intelligence expert. Cathye and I talk about all things Business Intelligence or BI. We talk about what BI is, and the skills students need to pursue a career in BI.

--------------------------------------------------------

Episode Transcript: 00;00;00;00 - 00;00;34;22 Welcome to the Oracle Academy Tech Chat. This podcast provides educators and students in-depth discussions with thought leaders around computer science, cloud technologies and software design to help students on their journey to becoming industry ready technology leaders. Of the Future. Let's get started. Welcome to Oracle Academy Tech Chat, where we discuss how Oracle Academy helps prepare our next generation's workforce. 00;00;34;23 - 00;01;02;10 I'm your host here appears in this episode. I'm joined by Kathy Pendley, an Oracle AI's director and business intelligence expert. Kathy and I talk about all things business, intelligence or buy. We talk about what the AI is and the skills students need to prove and see. You'll see where I get messed up and I start over. So in this episode, I'm joined by Kathy Penley, an oracle AI's director and business intelligence expert. 00;01;02;13 - 00;01;29;01 Kathy and I talk about all things business, intelligence, RBI. We talk about what be AI is and the skills students need to pursue a career in by. A little bit about my guest. Kathy is a business intelligence program director at Roseann and has 30 years of experience working with Business Intelligence analytics technologies. She brings strong project management skills and a clear methodology focus to each project. 00;01;29;04 - 00;02;02;05 Kathy has rounded experience in all areas of business, intelligence and analytics, including product project management. Sara backed up their project management of analytics projects to valuations and selection for business intelligence tools, analysis, design, development and implementation of analytics solutions. And she has developed both large and small analytic application patterns and systems. Welcome, Kathy. So to start off, can you please tell me a bit about your background and your job role? 00;02;02;07 - 00;02;41;05 I am a 1992 graduate of the University of North Texas. I have a B.A. in business computer systems, and it's very similar to what most colleges would call a BBA and management information systems. I currently am a business intelligence program manager at ROSENSCHEIN, and I've been there for about a year and a half. I focused at and is to understand the nature of our business and the latest technology and then determine how the technology can best assist our businesses and make informed decisions. 00;02;41;07 - 00;03;09;12 My professional career has been focused on business intelligence. Some call it decision support back 30 years ago are you might call it also analytics today. So it has many different names, but it has been in the business intelligence area. I work for Rosatom, which is an electrical contracting company. The majority of my career it's been in consulting, focused on analytics and beer across many industries. 00;03;09;12 - 00;03;40;11 So I have experience in many industries. I would say all but about four years of my professional career has been in consulting. That is quite background. You are an expert in business intelligence. So to start off, can you give me a high level overview of business intelligence or buy business intelligence and B, I can be thought of as a superpower of turning data into actionable insights that drive better business decisions. 00;03;40;13 - 00;04;06;23 It's not just about collecting and storing information, but analyzing it in a meaningful way to understand your business performance, identify trends and make informed choices. The steps to do that. The first step in be AI is to understand your business needs. You need to know what is important to the business, and then you can start gathering the data. 00;04;06;25 - 00;04;32;09 So what kind of data do you get and buy? You can have what we call internal or external data. Internal data is something that is within your company. Like sales could be payroll, could be h.r. An external data is something that you're getting external from your company. That's like social media, maybe even weather data. And then there are also different types of data. 00;04;32;11 - 00;04;57;07 You have a structured data that sits within a database, and that's something that you'll hear where you put them in tables and you join your tables together. But then there's also unstructured data, and that's like text documents, emails. Those are kind of some unstructured data where it can be in any type of format. Now you can do your analysis in an Excel spreadsheet. 00;04;57;09 - 00;05;23;21 And and that's okay for small individual type analytics, but for more complex enterprise wide analytics, something that you're going to push out to your entire company, it's best to create a model. And a common model that is used is a star schema. And all star schema is is just some tables joined together and you have what you call a fact table. 00;05;23;23 - 00;05;45;20 A fact table is nothing more than something that has a fat sales productive. What what is the key metric that you're looking at? Then you have your dimension tables and that's basically how you want to break out your data. So you're going to break it out by time or by location. You have a dimension table for each one of those with the attributes by those dimensions. 00;05;45;23 - 00;06;10;17 A good way to think about a dimension tables. If you're looking at something and you all look at sales and you want to see it by somebody says, I want to see it by product, by time, by location. Anything after the buy is going to be a dimension, a location dimension, a time dimension, a product dimension. So that's how you kind of build a model amongst the model is designed and built. 00;06;10;19 - 00;06;34;02 You need to then load it with data and put data into the model, and that is called data preparation. Some people call it ETL, some people call it BLT, but basically that's where you go in and you clean the data, get it organized, and you loaded into the model. This can be a long process. Once it's in the model, now you can start doing your data analysis. 00;06;34;04 - 00;07;02;05 There's various tools and techniques for years to analyze the data once it's in the model. This kind of generating report, creating dashboards, performing calculations are using data visualization techniques like charts and graphs. You get that built. Then you can start looking at the actionable insights. This this is where you have the analysis. It reveals pattern trends, hidden information that helps business understand what's working and what's not. 00;07;02;07 - 00;07;29;00 This knowledge is being translated into actionable recommendations that can be implemented and improve performance. As you look at this, we talk about building this and building the chart data visualization, don't underestimate it. There are classes. There are books. If you are going to be working with the users and working in building analysis, understand and learn a little bit about data visualization. 00;07;29;08 - 00;07;56;27 A quick note is people read left, right, top to bottom. So whenever somebody looks at a dashboard, the first place they look is in the upper left hand corner, their eyes drawn up there. So you would want to put your key metric in the upper left hand corner of the dashboard that makes it stands out that allows your executives to quickly get the information without having to spend too much time digging through tables to get it. 00;07;56;29 - 00;08;23;26 I had actually never looked at it that way. That's really insightful. I just had an moment thinking about the tables and graphs and charts that I built that I was really a wonderful nugget that you just gave. So now on to my next question. What are some of the different industries that are used by different industries for actually every industry and every department within industry uses by for example, you have your construction things. 00;08;23;26 - 00;08;54;02 COLONISTS All right, look out. I'll go a little, maybe a little bit more about install rates. Retail has sales and inventory, higher education. They're looking at enrollments and salting my look at staffing. But then even departments in these industries like your h.r. Might be looking at the retention of employees, and that would be across all industries. So there is pretty much within every organization, within every department, in every industry. 00;08;54;05 - 00;09;24;08 I'd like to go over a couple of examples of how it's used differently at a couple of organizations. I want to start with the construction installer right? This is a metric that many construction companies use to determine how long it's going to take to install a particular product. Say, for example, a conduit in our electrical contracting company, we have conduits and we have an estimated rate of say, 5.2, five feet per hour. 00;09;24;10 - 00;09;43;09 And that is saying that an individual, a worker should be able to install 5.5 feet of conduit in the hour. So what happens when a given project goes down to five feet per hour? And that's just a reduction of about 10%. Not horrible, right. 00;09;43;11 - 00;10;11;22 Talking about multiple projects that we have with hundreds of people working on these projects and this rate starts going down. If you had to put in 2.1 straight, the conduit at a rate of about $50 per hour, the cost alone at 5.2, five feet per hour is $20 million. If you go down to five feet, that could cost you $1,000,000. 00;10;11;25 - 00;10;35;01 Just in revenue. And then if you ever to take into consideration if you're if you're not installing as fast, you're going to have to put people on overtime to meet deadlines or you may have to pay a fee on your contracting. This million can easily double to $2 million. So it's important for our executives to look at the install rates and making sure that across our projects we're running it correctly. 00;10;35;07 - 00;10;56;24 But also it also filters all the way down to a foreman who might be managing a couple of employees or a couple of workers, because if they want to make sure they're doing their 5.5 feet per hour, if everyone does it at the lower level and at a higher level, we won't be making our money and we will be making our numbers. 00;10;56;26 - 00;11;16;18 And it's also important to look at this weekly so you can quickly address it and we will look at this in trends over time as well. So you can see how are we trending up or down on our list? All right. This is good analysis and this is some ways that our company uses analytics to help run our projects. 00;11;16;21 - 00;11;39;11 Now, another one that I think is more fun, and I did it. We had a customer, mine as several years ago did a prototype on this. And this is really very interesting and kind of fun. Look at the billboard. There are two main parts. You have a large sign that has the advertisement and the pole that actually holds that sign up. 00;11;39;13 - 00;11;58;06 So for thinking about hurricanes, if a hurricane hits the sign and it's attached to the pole, it's going to put pressure on the pole and possibly bring it down. If it's not a sign that it's not going to be as much pressure on the pole and it has a higher likelihood of staying up. So why do we care? 00;11;58;08 - 00;12;18;20 Unknown First off, there are ordinances that change over time. So a lot of times cities, counties are going to say, okay, you can't have a billboard up in this area anymore. It's become too crowded, can't have a map. But if you already have one, they'll let you keep it up. But if it goes down for any reason, you're not going to be able to put it back into the ground. 00;12;18;23 - 00;12;56;18 So that would be crucial. We would want that to go up. Also, what is the impact if we had that, if we have one that's completely down? What's the revenue impact? Is it one that's enough in that area that's visible, that brings in a lot of revenue, or is it one that doesn't bring in much revenue? So while it's expensive to take science down prior to our time, it might make sense to bring some kid some key signs down based on the fact that, hey, is there a new ordinance and are we going to be able to put it back up if the pole comes down? 00;12;56;20 - 00;13;18;07 The loss of revenue, what is the impact of the loss of revenue? Is it is it is it a sign that gives us a lot of revenue because it's on a major highway, or is it one it doesn't bring in as much revenue. What's the likelihood that it's going to get destroyed by a hurricane? So is it in the path of a hard time? 00;13;18;09 - 00;13;51;01 So internally, the company had data on the billboard, its location, and they were able to put it on an analytic map and an analytical tool on a map. They then downloaded the National Weather Service data for an upcoming storm. It's coming up National Weather Service data and it plotted out where the storm was going to be and the strength of the wind coming through that overlaid that where their signs were, where they had signs. 00;13;51;03 - 00;14;16;05 So then they could take it and say, okay, we're going to have extremely high winds going across ten of their sites. Two of them have an ordinance restrictions and they're not going to put the sign back up if it comes down. One of them is in a very highly revenue area, so they get quite a bit of revenue off that side. 00;14;16;05 - 00;14;50;06 So you want to make sure that sign comes down because the last thing you want to have happen is the pallet is going to the ground and not be able to put it back up and they are lost revenue forever. So these are some ways that people use it and are able to take the data from the billboards that they have, the attributes of the revenue, the ordinance that on all of that, on the billboard, and then take the external data of what's going to be the what's the likelihood that that that billboard is going to be taken down with the path that the storm is going to be going. 00;14;50;08 - 00;15;13;01 So those are two ways that you can two different ways that you can use analytics at different industries. I think that that is so interesting. Like, I never thought that you could actually use, you know, weather and RBI to predict weather patterns and billboards. I just think that is so interesting and a really unique way to use it and and very valuable, quite frankly. 00;15;13;03 - 00;15;35;05 Now, I'm going to move on to my next question. Now we kind of understand what buyers and we've seen a couple of examples. What are some job rules that are actually used by. I might go down two different paths and I'll use the I'm going to use the parking data, the Hurricane project example to kind of help submit talking about. 00;15;35;07 - 00;16;04;01 So you have first area is your business analyst or you're more of a business requirement gathering type. Some people might call it soft skills and those are the people that will go in and they will need to understand the technologies like like in this example, the graphing technologies, the mapping capabilities of of analytic tool can emerge different data. 00;16;04;01 - 00;16;28;06 And how easy is that going to be? So kind of get an understanding of the technology that you have. And then you also need to understand the business needs that the business wants to protect their key billboard drive. They want a high revenue Billboard said that if they go down, they're not going to be able to replace you want to make sure you focus on that and then you need to try. 00;16;28;09 - 00;16;59;06 So translate these requirements onto the technical resource team. So in this example, so what we had and I wasn't as involved in it, but you have a key analyst here that says, Hey, we have all this new technology. It's really cool. I knew that their business really was concerned about the billboard, so they took the knowledge they had worked with the business community and actually came up to design something and come up with these ideas. 00;16;59;08 - 00;17;24;08 They're usually the ones that manage the projects, keep timelines, make sure that they're not over promising or over delivering. And because people will come in with requests and they've got to be one that's and can politely say, we need, you know, this can't be done or this is going to delay the project and manage projects. And then you have to report and dashboard developers. 00;17;24;10 - 00;17;57;12 Unknown Many times it's the same person as analyst. They'll come in and they'll get the requirements and then a lot of times of build the analytics behind it. If I call it as a front end type scale, soft skills kind of analyst skills, and the project manager skills that are better is one way you can go the next crowd is more of a technical route and this person needs to understand how to build a data model to support the analysis so they'll be the one that will come in and say, Hey, we've got we got the data. 00;17;57;12 - 00;18;19;24 We want to know the wind speeds, expectations of this hurricane. So that's going to be kind of your fact. And you want to know where the wind's going to be the happiest. You're going to want to know about what time is going to hit. Right. So so is it going to hit around noon on Thursday or is it going to hit at midnight on Friday, about the time the location? 00;18;19;24 - 00;18;39;26 Of course, that's very important because, you know, when you had to deliver, where is it going to exactly going to be? And then, of course, you also have to have your billboard information in there as well, because that is telling you what is the contract type that you have. Why is a rental revenue? Are there any ordinances on this billboard? 00;18;39;26 - 00;19;05;26 So that's all your billboard information. So they're gathering all that information and they will build a model around that. And then there are also the ones who locate and claim the data, the ETL process, that they actually take the data, clean it, load it in there, will get they will go down and see where they can get the external data, like the data Weather Service, the National Weather Service, get the data and bringing them to the tables. 00;19;05;28 - 00;19;29;25 So that's more of a technical skills depending on the skill set of the analyst building the data model is where the two might meet, so the analyst could help build the data model. And also the technical person says we might see kind of an overlay of the two roles, but one of them is definitely more technical. And to be honest, you're usually the people that are more technical. 00;19;30;01 - 00;19;49;22 They will struggle with timelines. They really want to overdeliver because they see something super cool, Let's just do this so they're not as good with time management and not maybe not as good, but with people skills on it. It's important to have it there more. When you think of the back end and coding piece of it. So let's go actually go on to skill set. 00;19;49;22 - 00;20;11;00 So what are some of the skill sets that someone pursuing a career in the eyes should have? As you look at this, as you go down the different route, you may want to focus stronger in one area than the other. But I will tell you in every instance, you're going to need some type of analytic skills, some types of gathering information. 00;20;11;00 - 00;20;34;25 So if you have an opportunity to to work on team projects where you're gathering information and you're building some kind of right, you get requirements and you're building something that would be very helpful no matter which route you want to go. Of course, if you're going more into the analyst type role and getting the information, then you may want to take more classes in that focus more on that. 00;20;34;27 - 00;20;53;05 Definitely. If you're building the analytics, some visualization in there, that will be very helpful. In all cases, you want to kind of get a feeling for the industry. If you're going to be interviewing and meeting with companies, just get an understanding of the industry in which you're going to be applying for jobs or working in. That kind of helps you understand a little bit things. 00;20;53;05 - 00;21;12;28 You also need to have the technical skills to learn a couple of programing languages, and that's always helpful to couple. When I was in college, I will tell you I used neither one of them in my professional career ever, Although he might never use that programing language at your company, it does help you get into the mindset of what programing is about. 00;21;13;01 - 00;21;31;28 I know a lot. I would say I would take a skill class if you if you have the opportunity and I know a lot of a lot of big boardroom python now as well, but just get into a couple of good programing classes. If you're wanting to go down the more technical route, maybe take a few more programing languages and some more technical classes as well. 00;21;32;01 - 00;21;53;19 That is really, really helpful to kind of spread it out like there's two different pathways, but we need to know programing languages. I know through Oracle Academy we do offer sequel and and and database courses as well. So I do think that I'm glad that you mentioned sequel because that's, you know, one of the primary database languages. So now onto my final question. 00;21;53;21 - 00;22;14;00 If you could give a one piece of advice to faculty or students, what would it be? I can probably some this up in one word, and that is respect and this is respect of others and how you can do this various ways. But the key is one is time management. Be sure you make your meetings, your interviews on time. 00;22;14;06 - 00;22;36;07 Don't be late. I think in the world of texting and instant messaging, people go, I am running 5 minutes late and I might think that's okay. It really is. It because somebody has set their time aside to meet with you and now you're delaying them 5 minutes. So and which it makes meetings go over. So be sure you're on time. 00;22;36;12 - 00;23;06;11 I will tell you as an interviewer, if you are late to one of my interviews, I will probably have you come back two or three more times to make sure that was a one off case. So those are key lesson to others in the meeting. When you're in an interview, listen to others what they have to say. Repeat what they have to say, Make sure you understand it, get clarification, but make sure that you listen to them and then understand the needs of others. 00;23;06;18 - 00;23;25;18 So you may have you might be in an interview. I may be talking to someone, maybe a little bit frustrated. Maybe they have something going on, either personally and professionally in their life. But I understand their needs, understand what they've got going on, work with them and say, okay, I understand how I can help you. So focus on that. 00;23;25;21 - 00;23;46;17 To do this, you can start practicing this now in your personal life and it will just become a habit in your professional life. So if you worry, if you make sure if you got a meeting with a friend, make sure your you meet with that friend on time and it will become a habit. Secondly, be a motivator at is think of ways to bring your team up instead of down. 00;23;46;17 - 00;24;15;09 So it's really easy when somebody starts talking active thoughts. It's really easy to to spiral whole team down. But if you can find some positives out of it, that's really helpful. I know that there's something that I had a job and they were getting rid of the technology I had and I was experienced in. They wanted to keep me because I said I had the most motivated team in the company who cared about each other. 00;24;15;09 - 00;24;41;08 And we were always positive. So I can't underestimate how much motivation to have a positive attitude is important at companies. A big things to Cathy for talking to me about business intelligence or BI It's been really helpful and insightful to learn more about Oracle Academy and our resources. Visit Academy dot Oracle dot com and subscribe to our podcast. 00;24;41;08 - 00;24;54;23 Thanks for listening. That wraps up this episode. Thanks for listening and stay tuned for the next Oracle Academy Tech Chat podcast.