The Modern Data Warehouse
Hi i'm steve behebe. Welcome, you know thank you so much for joining us today in our first of a series of cloud platform summits, which are scheduled throughout the year we'll be bringing you, really relevant, topics, and use cases. That are especially. Pertinent. Uh during. This time. Today we'll be talking about how we build a modern data warehouse. And we also have some amazing. Customer, discussions, ahead. Now you know first i wanted to start with the context, of today, you know to best serve our customers, we're focused, and committed. To delivering the right outcomes, for them, for you. In every way it's about supporting. And enabling. This. Today our customers, are reimagining. And reinventing, their businesses, to lead them into a successful. Future. Now when we look at areas that are top of mind, they include, addressing. The economic. Impact. How do we help save costs how do we make sure we have a dynamic. Environment, where we can only, pay. For what we use, and maybe shut it off we're not using it. Ensuring, business continuity. And business, resiliency. Is obviously, key. Engaging, with employees, and customers, in new way with the rapidly, changing, dynamics. Of. Most marketplaces. How do we stay on top of that. And part of that is being able to rapidly, develop, deploy. Solutions. To market, whether that's customer engagement. Whether that's business, practices. Uh whether that's moving from a a, physical, mode to a virtual mode like we're coming to you today. Really helping to accelerate, digital, transformation. And finally and importantly. And specific, to the topic today. It's about informing, decisions, through predictive, insights. And again this will be the core topic, for today. Now next one of the foundational. Innovations. In core to our data management, and data warehouse, offering. Is the oracle autonomous, database. It is the world's only self-driving. Self-securing. And self-repairing. Database. By self-driving, we mean it automatically, provisions. Configures. Tunes and scales. From a self-securing. Perspective, it automatically, patches. Encrypts. Monitors, and remediates. Uh, so if it discovers, a security, action. It can automatically, help remediate. A specific. Potential, breach. And then finally from itself for a repairing, perspective, we offer, the same, level, of slas. Um and the lowest amount of downtime. Across, any cloud provider, we realize that as you move to cloud, you cannot, compromise. And so we've done just that and more with the autonomous.
So Next when you look at autonomous, database, it comes in, really, two. Key, features, or two services, that are available. The first is the autonomous, data warehouse, which provides the analytics, in seconds. It allows our customers to deploy, net new or move existing, data marks, data lakes, and data warehouses. To the cloud. And this will be our focus today. The second is autonomous, transaction. Processing, which is designed for high performance, transactional, workload. And that will be the topic, for, an upcoming, event. So now one of the other key, areas, of innovation, that we've done, is. Our focus, on choice of deployment. Where we can offer unique, access, to cloud innovation. Wherever. You choose. So whether that's in our public cloud, and that's autonomous, database, or our full, suite. Of data warehouse, solutions which we'll talk about. We can also take our entire, public cloud. And bring that to customer, prem, so there are many reasons, why customers, can't move to cloud. If it's data sovereignty. Data governance. Some sort of regulations. Uh, latency, issues or what have you highly sensitive, data that can't leave, the customer's. Premise, premises. We, can, offer a solution, you can get exact, copy of oracle's public cloud. In your own data center, and finally we also have a version, autonomous, database, on exadata, cloud customer. Where we can deliver just autonomous, database, capabilities. To your data center. And this level of choice. Um. Spans, our entire, data warehousing. Technology, suite as well. And that's what i wanted to cover next. So when you look at modern data warehouse, components. There's sort of five key areas, first is integration, where i need to be able to collect data. From multiple, sources, and those could be, heterogeneous. Sources, those could be multi-cloud. Sources, that could be heterogeneous. On primary, cloud. The important, piece is how do i pull it all together, in an easy mechanism. And as we talk you need the power of, the database. That's self-driving. Self-sufficient. And self-repairing. You need to be able to support, data lakes. Support. Advanced. Machine, learning based, analytics. And finally, data science, uh that's also, powered by machine learning, and uh andy will go into much more detail, in all of these areas. But, ultimately, our different, differentiators. Lie. And in, you know five key areas. It's a complete, integrated, solution, that meets all your data warehouse, needs. It's easy to start, it's easy to operate, and it's easy to analyze. You we offer consistent. High performance. You do not need to compromise. Performance. While moving to cloud. And in fact we can give you more. Secure. Environment. For your data warehousing, and analytics, with the comprehensive. Multi-layer. Security, approach. And finally as we discussed, choice of deployment.
Is Key. In terms of, the benefits. We provide to our customers. Now with respect, to those, customers. Um, we have many companies, around the world, of every size, representing. Every industry, who are making better decisions, and delivering, insights, faster. Leveraging our autonomous, data warehouse, technology, whether that's, spanning the likes of skye. Out front, taylormade. Cern. Adventist. Health. Um. Vodafone. Unicomer. And others. And you're going to hear from three, exciting. Customers. Today. So with that the agenda, for today. We will have a keynote. About, how we can modernize, without, compromise. Building the modern data warehouse, and that's going to be led by andy mendelsohn. Who's our executive, vice president of database, server technology, for oracle. That'll be followed by customer, spotlights. And finally a section on how we can get started, and what those next steps. Would be. So with that. I would like to introduce. Uh, andy, mendelsohn, andy thank you so much for taking the time. To, join us today, it's great to be here steve. Okay, so. Let's get started, here. So. Just to kick things off, um i thought it'd be interesting to put up this quote from uh. Eric brynjolfsson. Of mit. And. It's pretty it's it's very true. Everybody, has realized over the last few years. That. In almost every industry. Data. Is, the the most important, asset. Of the organization. And. Customers. Are very focused, on you know how can i leverage. This data. For business value. And. The problem is you know customers, for many years. Have been trying to do this. With some success. But, they've been running, into a number of issues, you know the biggest, issue i hear from customers, is that the data. Is all over the place, they have, data squirreled away in different lines of business. Locations. Eight central i.t. In the cloud, in object stores. And, for data analysts. One of the biggest problems is just finding the data that's relevant, for what. The business, tasks they're up to. And, and then, being able to analyze, it is is the next big step of course, and because the data scatter, all over the place. Data, governance, is really poor, security. Is, the weakest security, of all these different data locations. So security, is a big problem. Um, and because the data is so fragmented, it's hard to analyze it you need to try to consolidate. It, into a place where you can analyze that data. Through complex, etl, mechanisms. It's not something data analysts can do on their own they need a lot of help from lots of people, in the organization. To get to the data. To integrate it to analyze, it. And finally, you know there's been a lot of hype around machine learning. And i think. Most, enterprises. I talk to in other organizations, are very frustrated, that it's they haven't really. Been successful, in deploying. Machine learning algorithms. To get lots of value from their data. So at oracle, we have been working with our customers, for many years. On premises. To get value to their, data via data marts and data warehouses, we have tens of thousands of customers. Who are deploying, our. Database, technologies, our data. Integration, technologies. As. Well. And what we want to talk about here today, is. To help, both our existing customers. Who are already using, our, data. Warehouse, technologies. To modernize. Those infrastructures. Very very easily using. Our cloud, technologies. And also we will welcome, new customers, who want to move from, other platforms. That are not been. Meeting their requirements. Very quickly to a much more modern much more effective, data warehouse platform. And to do that we're going to talk about, four separate, areas. We're going to talk about. Number one our complete integrated solution. For the modern data warehouse. And then we'll go into a little more detail on, how to implement that solution, how you start how you operate, how you make the data secure. And finally how you you get value out of the data, via analytics. Okay, so we have a complete integrated, solution. Let's move on. Um so this is a general architectural. Picture, of, what a modern data warehouse platform. Needs to look like, of course on, the left side you've got all these data sources. For the data, your analysts want to get to. Everything from package, applications. To, third particle. And third-party, databases. Data, can be on premises, it can be in other clouds. You have, batch processing, to get data into your, data warehouse via etl. You have, events. Using. Kafka, usually to move. Streaming, data and you have tools like oracle goldengate, to get, change data capture, from, various data sources into your database, and finally data can be, in files, also it could be in object stores.
And Then you need a processing. Refinery, that we call the data refinery, to, to take that data, and transform, it into a. Form. That can be easily analyzed. You move it into. Either object stores, or persistent. Databases. Like oracle autonomous, data warehouse. You, form data likes using data in object stores. And and then finally you you apply the, analytics, and machine learning models. To get value. Out of that data. Some of the key components. Of the oracle. Modern data warehouse. Are. Listed below here. Steve quickly went through them i'll go them through them a little more detail but they basically map to the picture, we just, saw, you know the integration, technologies. Are pulling the data out of the various data sources. And oracle, has very powerful data integration, technologies, that we've been using, for with our customers for many years things like oracle data integrator. Golden gate. For removing, data. From. Various sources into your data warehouse, we also in our cloud of course have kafka, service now as well for doing that. Um, and then the, the central point of our our data warehouse, platform, of course is the oracle autonomous, data warehouse. Steve talked a little bit about that and we'll go into that more detail, later in the presentation. But this is not only the platform. For doing high performance, parallel, querying, of your data in the data warehouse, it also can reach out, transparently. To data that's in your, in, the object stores, that are forming. The the data like, if you're. Going in that direction, as well. And finally, you know you get to the analytics, side you can use traditional, analytic, tools. Like oracle analytic. Cloud, or third-party, tools like you know tableau, click etc. And then machine learning platforms. For, getting. Doing predictive, analytics, on your. Data. The autonomous, data warehouse, is a very, unique. Technology. That oracle, has brought to its cloud. Over the last couple years. And. The first. Bit of differentiation. Around autonomous data warehouse, is that it is a converged, database. It's not just a simple. Uh. Traditional, relational database, they can look at structured, data only. You can look at all your, data. In a single, data store, you can have both traditional, relational, structure, data. You can have. Newer data types semi-structured, data types like json, xml. You can have graph data, you can have, spatial, data. You can have multi-dimensional. Data. And, we have a powerful, sql. Engine, that lets you, do. Queries, across, all. All the different data. All, from one place, okay, no one else can do that. This converged. Engine, can do any kind of data management, activities. Relevant, for your data warehouse. You can do powerful, sql analytics, you can do machine learning on the data right in the database, itself, we have algorithms, we've been working on, for 15, 20 years, that, have all the popular, modern. Machine learning algorithms. We have in-memory, columnar, technology, for doing, very, powerful, very high performance, analytics. Uh we can do in internet of things. Streaming. Uh blockchain. Right in the database, itself you don't have to, create go to another specialty, database, engine. To do blockchain, or any of these other technologies. And finally this powerful sql engine not only works against, querying, data in the database, itself. But it can reach out to data that's in files, in the object store, so you can have, single, sql queries.
That, Join data in. Files and object stores, in various formats like parfait. Um. Json, etc. With data that's in tables, in in, the, autonomous, data warehouse, itself. And finally, the big advantage, here is, you know to get your job done in other clouds, you have to use lots of different engines, if you want to do things like graph, and spatial, blockchain. Json, document. Analytics. Here, everything is converged into one. Database. So you can have a very simple, method. To, query, and analyze, all your data. And as steve mentioned, earlier, we, have very, unique, deployment, choices, at oracle. Of course you can run autonomous, data warehouse. And our solution, for the modern data warehouse. In your, in our public cloud data centers. But, you can also. Run, the same technology. In your data center either by having, essentially a private cloud of autonomous data warehouse, technologies. In your data center itself, or you can put our entire. Gen 2 public cloud infrastructure. In a dedicated, region. Dedicated. Just, for your business. Okay, so that's sort of a high level of our complete solution. So let's, get started now so what do you do it's. To to provision. A, an autonomous, data warehouse. It's really, easy, you just go to your web browser. You do a few clicks where you say how many, what's your, your best guess for how much compute, and how much storage you're going to need, you know give us a password. And you press a button. And in a couple minutes you get the world's most powerful. Database, technology. Completely, provisioned, for you. And then you go to the next screen and you can start loading data. Very simply. Via, our online, tools. And then, as you your needs get more complicated. And you want to put together, a, more. Complex, solution around autonomous, data warehouse, we offer, you. At the push of a button. Provisioning. Of much more complex, infrastructures. For dealing with departmental. Data warehouses. Or data marts, enterprise, data warehouses. Data warehouses. That are integrated, with data lakes, on our object store, doing machine learning. And. All these things can be provisioned, at the push of a button, and we provide, you, with, collateral, to help understand, how to use these. Technologies. As well. So, for those of you who are existing, oracle data warehouse customers, on premises. And you're looking to modernize, to a more, modern, cloud, architecture. Going to oracle, solution, is the easiest. By far. Tech way to go. You don't have to make any changes to your existing. Etl. Scripts. Your bi tools, all work the same security, policies, can work. All you need to do is. Repoint, the etl. To our data warehouse on the cloud. You take your tool existing, analytic tools, you re-point, them to autonomous, data warehouse. And you're done, that's it, you know using any other kind of. Cloud, technology, to modernize, your data warehouse, is a much much more complex, project, and we have a quote here from, a healthcare. Provider, who who looked at the two. It's just you know, simple repoint. Versus a complete re-architecture. Of everything, you've done rewrite all your sql queries all your etl, scripts, everything. So, for those of you who are existing, oracle data work house customers, it's sort of a no-brainer. Look to oracle's. Modern data warehouse, solution. Those of you who are. Existing, oracle fusion applications. Users using things like fusion erp. Fusion hcm. We offer you a complete, solution, out of the box. For your data warehouse, no mess no fuss. Where we do. You know, complete, etl, for you from the data from your, data sources, in your fusion applications. We do. Pre-packaged. Analytics, for you. And dashboards. Etc. And so. And then you can, you know grow from there you can once you get what we give you you can say hey i need more data from other data sources. And you can easily, use autonomous, data warehouse, directly, to add new data sources. To add more queries, etc. So it's a fully extensible. And customizable. Platform. So again if you're a customer in this space it's the easiest way to go to create a modern, data warehouse. So let's look at the third area you know so once you've got your data warehouse provisioned. What do you have to do. Well autonomous, data warehouse, is a very unique.
Technology. Out there in the industry. Not only do we automate the provisioning. And configuration. Of your data warehouse, like all the other cloud vendors do, we have unique. Tuning capabilities. So that, we will automatically, do self-tuning. So, your data analysts. Effectively. Are. Dbas. Themselves. They don't really need to know anything, about the tuning, they don't need, another, dba, to come in and help tune, the system for, them. We uniquely, automate, the security, for them. You know there's constant, security, patching, needed, in this world in this day and age. And we do that all, for you online. With no downtime. No one else can do that. We uniquely do elastic, scaling and i'll go into more detail, what i mean by that as well. And the the key goal here is to automate, everything, around, managing the database. So your analysts. Can, basically, do self-service. They don't need, you know the the. Dba, anymore, to provision. And, tune, anything, for them. And it frees up the dbas. And analysts, and developers. To, do more innovation, and create more value from the data which is of course what the, business. Really, really, values. Um. Oracle, autonomous, data warehouse has very unique, elastic. Scaling, capabilities. The other vendors, talk about elasticity. But it doesn't apply to their databases. Competing, platforms. Like redshift, and snowflake. Are have very traditional. Shape, based, provisioning. Technologies. Where, you. Choose from one of a different. Various shapes so for example. You might, get a shape that has eight cpus, and so much storage and memory. And when that runs out then you go to their next biggest shape which might be 16, cpus. And more storage, in memory. And when that runs out you have to go to even bigger shape that might be 32, cpus. With oracle autonomous. Data warehouse. If you were wrong with your initial provisioning, and, let's say eight cpus, wasn't quite enough but 10 will do it. No problem, you just add a cpu, you go back to the console, all online, you add a cpu. And you're done. No problem no mess no fuss, no wasted, cpus. That you incur, when you go to these shape, based, provisioning, technologies. The other thing we do that's really unique, is. Sometimes, people have spiky, workloads, you know normally. 816. Cpus, is good enough. But, occasionally, there's this batch workload, that runs at the end of the day or the end of the quarter, that needs, maybe 2x, or even 3x. The compute workload to get the job done in a timely fashion. With autonomous, data warehouse, auto scale. You give us what you think is the baseline. Say eight cpu is what you need, and then we will automatically. Scale up up to three times that baseline. So if your baseline, is eight, we'll scale up to 24, cpus, automatically, for you, on demand, based on your workload. No mess no fuss you don't have to do anything, yourself. And then we also, let you. Know of course manually, if you know your business. Is going to do some huge processing, run at the end of the quarter that needs 40, cpus. You can go in there on the console, just. Rev things up from 8 cpus, to 40. You know run temporarily, with that. And then, go back down, when you back. When that that workload, is over, so the bottom line is elastic, auto scaling, drives, down costs tremendously. Beyond what the other, vendors, are able to do. Um, we you know we mentioned earlier we have automated. A, self-driving. Self-tuning. Database, that's what autonomous, database, is all about. Uh we automatically, do the the optimization. For you by doing things like creating, uh summary, indexes, for you, uh. We automatically, take advantage, of things like in memory kilometer. Technology. For you, we automatically, paralyze, all the queries. Select the best plans, etc. All done via. Artificial, intelligence, algorithms, of various, sorts including machine learning, to give you the best performance. And then. The autonomous, data warehouse for those of you already. Using oracle technology, on-prem, of course. Uses. The exudated, technology, that. You are probably using on-prem for your data warehouses. We transparently. Are using it under the covers and taking, advantage of all the, great, performance, innovations. That are in exadata. All transparently. To you. So. We have a customer here vodafone, that i'm sure a lot of you have heard of they're a global, telco.
Provider. And, they were existing, oracle, data warehouse, customer, on premises. And their example, of a customer who, was very, able to very easily, modernize, their data warehouse. By moving to our cloud in autonomous, data warehouse. And as part of that move. They got, six times better performance. They're taking advantage of the auto scaling. Capability, i mentioned, earlier. To. Automatically, grow their warehouse. And at very very low. Cost. Another, unique. Capability. Of autonomous data warehouse, of course is that it's based on the oracle database, security, technology. This is by far the most mature. Powerful, security. For, data, data management, in the industry. Um. And, we automate, everything that's possible, to automate, automatically. So for example, um encryption, of course all the data is automatically, encryption. Encrypted, you don't have to do anything, take advantage of that. We have a technology, called database, vault that makes sure that only you can see the data, in your data warehouse. People from oracle, are completely. You know locked out from that. And then, as i mentioned earlier. You know if there are some security, updates that we need to apply for you we do that all online. With no downtime, again a very unique, capability. And finally we have a tool called data save. And data safe. Is a tool that lets you do things that we can't do automatically. So for example, you have to manage. The users of your data warehouse, and the privileges, you give them. And then we will do risk assessments. So we'll actually look at, what access. In, the real actual data where else your users are, doing you know what queries are they performing what, tables are they looking at, and we make sure, you haven't given them privileges, to do more than they really need. We do automatically, auditing of all the data for you, and put it into a central repository. And let you run analytics, against, the audit trails. And, we also automatically, will help you find out, that you might have some sensitive, data that you didn't know about and you might want to mask that data, when it goes into a, test environment. And finally, all this powerful, security. That's in the data warehouse. Database. Can be extended, transparently. Out to data that's in files in your object store, so you can apply all the same powerful, oracle database. Security, mechanisms. To data, in your, object store. Okay finally, you've got all this data. Now you want to analyze, it. Just as a free part of the service we provide, some visualization. Tools for you, and we provide a very powerful, tool called, oracle, application, express, or apex. That lets your data analysts, who don't want to write a lot of code. Build simple, applications. Really fast. At really high productivity. So they again can, can be self-sufficient. You want your data analysts not to need anybody. To provision the database for it for them to tune it for them. And finally to write code for them they can do everything themselves, with oracle's. Modern data warehouse platform. Oracle. Of course on our cloud provides, a very, powerful, analytics, tool called oracle analytics, cloud. For those of you who are using oracle bie. On premise. It there's a very simple, path from bie, to oracle analytics, cloud which is basically a, big superset, of what you were using on premises. And it's a modern, tool that does everything, from, you know traditional. Governed, analytics. To self-service. High, powerful visualization. Of all your data using. Machine learning, as well. And then of course oracle. Knows that all of you. Used, lots of different partners. For your analytics. For your, etl. And we work with all those partners, to certify.
That Their technologies, work well, with our modern. Autonomous, data warehouse, platform. And finally there are the data scientists, out there. And for you we have two, choices. A lot of data scientists. Like, taking, data. You know out of their databases. Or files, and, and. Working on them in a separate repository. Using open source machine learning algorithms. To build their. Data models. We have, oracle cloud infrastructure, data science, for doing that sort of thing. And, for those of you who'd want, more simplicity. And you want to build machine learning algorithms, on data that's right, in the autonomous, data warehouse itself. Or in the object stores. That are accessible, at from the data warehouse platform. We have machine learning algorithms, that we've been working on for years. That are built right into the autonomous, data warehouse itself. So without any mess or fuss no etl. You can build very powerful, machine learning models. That you can then deploy. You know. At your, pleasure, where, where they're needed. And finally both of these. Data science, platforms. Support, auto ml. So this is a very sophisticated. Technology, that we've built, in our research, labs at oracle. That lets, any data analysts, become, a sophisticated, data scientist. You just show us, your example, data set. You, show us the data. And the various features, we you want us to use to to, do the predictive, analytics, you'd show us the result. Of of, for each row essentially in a table that what you want the prediction, to be. And we will then go through our catalog, of machine learning algorithms. And figure out which is the best algorithm, for you, and, if there's more than one possible we'll give you a couple of choices. And then you can just, pick out the machine learning how model you you want to use, and go from there, very very easy, every, data analyst in your organization. Can now, do what, before, only sophisticated. Data scientists, could do. And of course as i mentioned earlier. Autonomous, data warehouse. Has various, unique technologies. That deliver. Very, high performance, using, very few, uh, compute cpus. And therefore, deliver, very low cost because on the cloud you just pay for what you use. And so when you compare, autonomous, data warehouse, versus, aws. Redshift. Or versus snowflake. You will see, that we use, far less compute to get the work done, and therefore, save you money. Plus i mentioned earlier we have, our, fully elastic. Auto scaling capability. Aws. And snowflake, are first generation. Shape, based provisioning, technologies. They, they require you to pay for lots of unused. Cpu. We have auto scaling, so you pay for just the cpus, you need, and we'll automatically, scale up just for a few hours or a few minutes. To get your job done. And no one else can do that they require you to move to these bigger and bigger shapes with lots of wasted. Compute. So, we. Okay. Um, just as a final thing i mentioned, i talked a lot about, you know how we have lots of existing data warehouse customers, to, can easily, modernize. Their infrastructure, by moving to our, autonomous, data warehouse and other modern cloud. Data warehouse platform, technologies. But we also have customers. Moving from other places. Um, so for example, here we have ripley's, a leading retailer, credit card bank. They were using. Aws, redshift. They, moved to autonomous, data warehouse. They lowered their cost 25. By using the techniques we mentioned, earlier. They're getting better performance, on fewer, compute. And. Having a great time using, oracle's, modern data warehouse platform. And just to summarize. You know, what are the unique, differentiators. Of oracle's. Modern data warehouse. Cloud platform, versus what you get on aws, and from snowflake. Number one, we are using a converged. Database. Technology. All. Your data types, from structure, to graph. To. You know json, to xml. Can be analyzed, with a powerful. Converged, sql, engine. In one place you don't have to fragment, your data across lots of different, specialty, databases. To be able to analyze. Your data. We have the world's most powerful, data management, security. Um. That. Lets you secure your data both in the oracle database. Itself. But also, it can be extended out to data, that's in, files in your object store. We have of course self-drive. Self-driving. Database, very unique. Self-tuning. Platform. No one else has that. We talked a lot about our truly elastic. Auto scaling, capability. No one else has that. And finally, oracle, uniquely. Lets you deploy our cloud technologies, both in our public cloud, and in your data center as well, you know amazon, snowflake, can't do that either.
So Just to summarize, we talked about. Oracle's modern data warehouse, platform. Gives you a complete, integrated, solution. For modernizing. Your data warehouse. Technologies. And we went through the different steps, of how easy it is to start. How easy it is to, load data. Operate, and secure. Your modern data warehouse. And then we talked a lot about our powerful. Analytics, and machine learning technologies, both from oracle. And from third parties. That integrate with our, our platform. And with that let's move back to steve dahid. Great andy, thank you so much you know what a. Excellent, overview. Great insight, and what a great way to kick off this series of virtual events, thank you so much. Now i think we get to my favorite part, you know i'm getting my popcorn, ready, it's time for our customer, spotlight. We will showcase. Uh. Different use cases, spanning, departmental. Enterprise. And data lake use cases, so. Please, um. Welcome, our customers. Hey thanks steve everybody scott wiesner, senior manager product management. So look throughout our history, our purpose. Has always been, to help our customers, with better data management solutions. Because whether it's optimizing, ad sales. Improving health care, or unlocking, nature's secrets. Your ability to collect. Analyze. And secure, data, is your key. Competitive. Advantage. So in this segment we're really honored and i'm actually really excited too, to have three customers, who are leveraging our platform, to contain, costs. Power innovation. And then of course mitigate, risk that's a big one right. So, this is a live event. But our first guest was actually very successful, so he was actually able to go on vacation, this week. We previously, recorded, my interview with derrick hayden a couple weeks ago, let's play that video. We are joined by derek hayden who's the senior vice president of data strategy, analytics, for out front media, derek, welcome to the summit. Thanks scott thank you for inviting me, absolutely. Thanks for spending a little bit of time with us today. So can you tell us a little bit about outfront media. Sure, outfront, media is one of the largest. Out of home media, advertising. Companies in north america. We have. Over 500, 000. Advertising. Displays, in the us, and, canada. And. We. Deliver, brand impressions. On behalf of our advertisers. To their customers, through, a combination, of. Technology. Location. And creative, services, so we think we have a really nice. Asset portfolio. And hopefully, when you guys are out and about whether it be driving around. Or, on transit. Uh you'll, see some of our assets, in front of you. I've definitely seen the billboards. Uh, love to see the logo. And. Again thanks for having thanks for coming on the event here today. So, let's go back in time a little bit, um what was the situation, about six years ago, before you had your current system in place kind of give us go back in time a little bit tell us kind of what was going on at that time. Sure. About six years ago we were a division, of cbs, corporation. So. Um, and we were certainly not one of the larger ones. And so, our charge was to make our revenue numbers and keep our expenses, down and that was you know essentially. What we, focus on. Um. We were spun off, as a as our own company. And. All of a sudden, um. You know see we had a c-level, suite that was asking, real business, questions. Um, we had, investors, that we had to. Uh give answers to. And. A lot of our information, even though we were, uh fortunate, to bring all that data with us i think that was very fortunate, as a new company. To have 15. 20 years worth of data, at our disposal. Um, everything was very siloed, it was uh it was really hard. To get consistent, answers. Um, the it, group, at the time was constantly. Running. You know running around trying to give answers. Um. We weren't cross-functional. In that. At that time. And then, they would have kind of savvy, users. Uh on, sales and finance, teams, who would. You know get data from different sources. And aggregate, it. Uh. And it became. It became. Gospel, so to speak. But. People weren't. Speaking the same language, necessarily. They weren't looking at things the same way, and it became problematic. To say that you know the answer that people were giving. Was a governed. Answer that we all trusted. So putting down the context a little bit so. Spin-off. You know you i think you recognize, at that point that, the there was a bunch of, siloed, information, people were kind of left to their own devices, there was no governed, access. Uh no consistency. I would imagine. Um. You know what what was, what was the kind of the mindset, there, because i think you stepped in i don't want to. Cover that. You realize, right away. That this was not sustainable, right what happened next, yeah we were we were in a meeting with the cfo, and. The sales strategy, group, and we were kind of kicking around. You know how we could.
Uh Get information, to people faster, the sales strategy, group had started. Creating an access database. And they were spending a lot of their time. Wrangling, data. Manipulating. Data. And then. Distributing, data so, um. You know the cfo, kind of said you know we need to we need you guys to work on strategy. You know not. Data. And um. We didn't have a date we didn't. When we left cbs, there was no analytics. Practice, or platform, for us. So. He kind of said we need something like a warehouse, or something, some sort of david warehouse. Uh, that was his that was his uh technical, term. And so um. You know i kind of raised my hand said you know i'd be willing to take that on, um. I was kind of on the finance. And real estate application, side. Um, we were. We were in oracle, technologies. My group. Um, and i said you know let's. I'll take that on because i just felt like as a. New company. Our. Services, as an ebs. Support, group. Could be easily outsourced, so i kind of saw it as an opportunity. To. Maybe make us more. Attractive. Long term. As as resources. So the cfo. Recognized. That, okay, i know the power of data i know the value of that data. But we have the sales people trying to wrangle that data that's not what we hired them to do we, needed them to be selling. So you volunteered, good for you, you weighed right into this and said. You know i i can actually set up a data warehouse i got some background, here so so, so that's kind of the context, i think where you started and i would imagine our other customers have a similar situation, even whether it's a spin-off, or a. Line of business, or, we all face the same things where. We want to be data providers, to the business and let them do what they need to do, versus. Having them to do everything, so kind of what happened next after you ever after there was a realization, from your cfo, of all places. The data warehouse, well i mean ironically, when we were spun off we actually didn't have a data center, to go along with us so he said we want this warehouse, and, like to, do something, and, you know analytics. And we want to do it quickly, so. We evaluated, a bunch of solutions, and, especially at the time. That work will have, had cloud database. And vi cloud service. Um it seemed. And we had oracle, tech, technology. Uh expertise, as well so it seemed like the best. Course to success. Was, was to adopt, a cloud-first. Platform. Oracle technologies. And. You know kind of, address. The sales, groups. Needs. Over a period of time so. We kind of laid out a plan. Of about a three year plan, to onboard. 700. Something users. For the sales organization. Uh. And everyone agreed on it, and then we. Kind of got into it and. We got the, five or six, key. Measurements, from the sales group that, they wanted. To be. Distributed. And. Not only, within, three months not only did we have those, but we had another, 10. And we had. About 250. Users, on, right away. And then. The finance, and real estate. Teams were kind of hearing about this and, started to you know peer over the shoulder, like what are you guys doing over there, hey we want in on that so. Uh we really. We fast tracked a lot of this stuff in the first year we had everybody up, you know.
That We had planned on for three years and we were covering, three lines of business so to speak. Wow. From my group's, support standpoint. Wow nice job i'm glad you volunteered, for that that role. Um, so, so so real quick, um. So i think there's a critical part and we'll come back to kind of what we lessons learned uh towards the end here but, you started with the sales team which was good and you built that plan so it sounds like the first thing was let's build data warehouse. Um let's provide dashboards, for the sales team. Um, there's a little bit of destruction, there right on looking at right around 2017, what kind of what happened there yeah we were the. In that 2016. Time period we were, feeling. Really good about ourselves, and delivering. As quickly as we could possibly. Get requests, in, and, you know we were adding value and i feel like we're streamlining. You know the reporting, side and, really starting to introduce, visualization. And, analytics. As a, as a concept, in the organization. As a technology, team we kind of got you know caught up in, in all the accolades. And things kind of went haywire, so, we started running into, issues, with, space, in our for our data, we had to, cloud database, we had to pick, you know a economical. Cpu. Range to stay in, and, so we were, you know we kind of got ourselves, turned around so we just started prioritizing. What do we really need to do we had to pump the brakes in 2017. And, kind of completely rewrite, our warehouse. So, you know. That helped us address some space issues, it helped us address performance. But, it took about six months out of. Our production. Of new content, so to speak. Um, and really. You know kind of. Sidelined, us for that period of time. But then we, kind of got right back into it, you know we did an analytics, assessment, we wanted to make sure we were on the right track. And, we didn't want to get into that. That space again. And, that's about, when adw. Kind of materialized. In the oracle portfolio. So, we started poking around with adw. And, it addressed a lot of the things that we were struggling, with, we did the universal, credit conversion. Uh, kickoff, 2019. Which kind of opened up the doors, to all sorts of new tools. It allowed us to address, some third-party. Data that we've never really bought into our. Warehouse, environment. And. Uh we brought that in through adw. And then that's a whole nother interesting, story. Somewhere. Um. That you know we were able to turn to turn around the media, spend. Analytics. In, like three days, literally. Um, at least you know get it out, back out into the field in like three days. Um, so and then db came out on oac. And odi. And spatial studio, so, we've really. Kind of, opened up our world now by, by uh. Kind of. Sticking to the oracle cloud, platform. Wow, thank you so, uh. Interesting, so recognize, the need for kind of centralized. Uh. Data management, if you will. And then you realize, okay we probably better off in the cloud, start doing things manually. And then here's this autonomous, thing which seems to be the right thing to do so so let's actually double click into your architecture, a little bit. Um tell walk us through. Kind of what the data sources are i'm sure our audience wants to know kind of what does this look like maybe at high level what your data flow looks like and kind of go from a service level description, here, yeah so it's probably not, that.
Unfamiliar, To people, we have what we consider, on-premise. Data sources, which would be. Oracle ebs. We have a db2. I mean they're both act they're actually in separate hosted data centers they're actually not, anywhere near each other, um. We have our infamous, access. Database, that is still. Somewhat. Alive, i'm trying to kill that off but um, and of course excel. Always, creeps in as, a source or, or, probably. More recently, smartsheet, my new favorite thing. And then we have some cloud sources, uh we've worked with cloud sources, we've got boosters, a is a, media, centric crm, tool that we use, we have some ibm, cloud. Uh kicking out there and then this, third-party, media spin, that we consume. Right now is is, another, huge source so like our main sources of data there on the left. And it's kind of a hybrid, you know i mean it's kind of coming from every direction. Yeah um, so, we as a. Again as a technology, team we kind of wrangle it you know we use odi. Little pl sql tricks. Things like that but um you know we're really looking to to odi. Um. As our. I call it data orchestration. Method, methodology. And then our warehouse. Is, you know, adw. And there's tools off of that which we actually use apex. And spatial, studio. And, um so that's been great. And then the visualization. Side, is oac. Perfect. Not uncommon. I'm sure a lot of folks in our audience. Have a lot of different data sources, and probably. Want to add even more data sources there's just a. A hunger for all the different data from from various, sources. So tell us a little bit about, uh why you chose oracle, for this uh, for this data warehouse you're basically. Implementing, something new in the cloud and, what what do we get for you, yeah i mean it was it was. We were an interesting, position, because we didn't have any preconceived, notions, of what. A warehouse. Or an analytics. Delivery. Should be. We went through a very. Long. Practical. Evaluation. Of technologies. At the time. And again, we actually had a partner. Come and they recommended, microsoft. On our behalf, and i i, said if i'm going to own this i i don't believe, that that's where we should be. Um. You know we had oracle technologists. We. You know we've had a long relationship, with oracle. And oracle, technologies. Um. You know my team was, uh, ready to take the challenge, you know. They were ready to move into a different, role, in the company. And. You know we were able to kind of repurpose. Existing, resources. Into. The analytics, space. So it really didn't cost, anything. Uh head count wise, to. Take this on in the oracle platform. Um, so it kind of was a no-brainer. There you know the cloud, was perfect for us it let us kind of, get everything going quickly, and i think the time to delivery, from when we. You know kind of spun things up to actually start delivering, was was so quick. Um. That. You know. There was no chance for second guessing. Um, there was, you know. We saw immediate, success, and immediate, value and really we never looked back. Since that time i mean it's been six years, and you know the portfolio, that oracle, has. Put out in the in the cloud analytics, space. Is. Really impressive, and i've, seen it grow i mean there's been changes to how you, you know. Get to it but i mean right now i mean it's, it's a. Really powerful. Um. Cloud. Offering. Um, and we do everything through. Perfect yeah i think you think first of all thank you for uh sticking with us. I do believe we have a lot to offer and you're proof of that and we'll hear later from some other folks as well. Tell us a little bit about, um what your experience. Let's talk about just the data warehouse, itself. As you move from kind of doing things manually. To. Leveraging the power of autonomous, data warehouse tell us a little about your experience, there. Yeah so i mean we were, you know. Building the warehouse, out in cloud database, and that was great, um. Obviously, it's, foundational. To our, uh, our success, story. Is, is how quickly we can we, created the warehouse. Yeah, three years, instead of three years fifteen months uh yeah. That's the market, that's pretty amazing. But you know again my, we didn't have a data. We didn't have a data center we didn't bring any, dva's. With us when we when, we were spun off so, um, know we had true technologists. Um. And, i really wanted them concentrating. On data, um, and so. In order for us to. You know keep this going you know not only do we have the data warehouse, relay but we had to worry about space, issues. Um, you know backups. You know patching. Security. We were limited, or hampered, i would say, by. We weren't able to scale up and back. On the cpu, side you know we had to commit if we wanted to move to that next level. So. You know we really, i mean. Those first three years we were feeling our way around, and. We got we had such great success, the, the last part of it was. Trying to get my technology, team. To focus on data.
And Partnering, with the business. Um because that was the most powerful. Part of our offering. Um, was being a partner. And. I didn't need them to spend. You know 30, of their time. Dealing with the infrastructure. Even on cloud, you know even on just the technology. So that's really the with the push. Obviously, for adw. You know we get. Self-patching. Self-repair. It scales. We don't have any uh you don't have any issues. With storage. So, it kind of just you know, allows, us to have that much more time back on the technology, side. And again just keep producing. And. And when we partner, with our. Our customers, and. Say this you know now we're not even an i.t, group. We're a technology, services, group so we really. You know try and partner and i just tell them like don't. Think there's any limitations. Just tell me what you need. And. And we'll figure out you know, they don't, necessarily, want to care about the back. And and this infrastructure. Here allows us to say yeah we can do that. Yeah yeah that's perfect. You've never had to say no to them for anything. Oh that's nice, yeah. You definitely, you know as it professionals, we always want to be partners with the business, and delivering, what what they want you know yesterday. Instead of them trying to figure out on their own which is what happened initially. Um. One of the things for you know certainly a sales team is you know they need to get out sales quotes, like immediately. And some of these things can be really, uh complex, so. How does the platform, and particularly the data warehouse and how does that serve that need can give us an example of you know kind of you know like the query performance, basically. Yeah so there's a couple things there's there's one, finance, team is really interested, in, the sales directors, in, the revenue, forecasting. Back in the in the timeline, slide part of that data warehouse, rewrite, that we took. Was to really facilitate. Some of these revenue forecasting. Queries, they were complicated. And so. When we first got to abw. We said let's, see what this thing can do we just, we took the data we moved it over. We didn't. Didn't pull any levers or anything we just started writing queries. And, so the first time we ran our revenue, forecast. Inquiry. Uh it took about six minutes. And i sat there i was like oh boy this is, this is not good. But the second time we were in it it took two seconds. Wow, and then, the more we ran. The better the performance. And you know to, to, be fair, the funny i'm not a technology, i'm not a. Coder, or not a, technologist. Per se, you know. My team would say well we can make it faster than that you know, sub two. Closer to one like we can make it faster. And i said yeah but we made it faster in about you know, two minutes, like. We took we had to stop and take six months to do it um, this is probably. The right way to go i mean most users, aren't going to, notice a difference between. You know. A highly tuned, um. Query and one that was, basically. Auto-tuned. Um, sure, yeah so, you know that was great. And then, the, you know the corollary, to that is this media spend data that we were bringing in from the third party was something that we weren't really familiar with, so not only was it a lot of, data. For us to ingest. But, we didn't want to spend a lot of time modeling, and indexing, and all that so we just let it run. And we kind of let people, hit, against, adw. Just as they normally would against the data. And so, we really haven't really touched. That. Data model and just let adw, do the work for us. Wow, that's amazing, so six minutes, two seconds i'd imagine it's getting faster, and you're right business doesn't care they just want as fast as they can. Whether it's auto-tuned, or manual, but you look time-to-market's. Everything, and so to speak you can do more with the, short amount of time that that's a great story uh appreciate, that, so uh let's kind of uh get out of the architecture, a little bit and tell us you know a little bit more about, you know some of the outcomes you have you know where you're going and, and i'm sure the audience wants to hear a little about hey along this journey what did you learn and how do we how do we help our peers. To go even faster. Um. Understanding, the mistakes that you know or, deviations. From your path that you had before so let's start with a little bit of that about the outcomes. Yeah i mean, we've touched on them when we talk but you know just to kind of reiterate, that we have we have a data warehouse, without a data center, oracle is our data center for for lack of a better word so it really allows us to, focus we've consolidated.
Our, Cross-platform. Reporting, on into one place. It's governed, from the standpoint, of you know people go there and it's trusted, if it doesn't come from there there's definitely, some, at this point some. Some questioning. Um, c level has embraced. You know. The data, and the visualization. Um, yeah i think i mentioned to. Chatting at one point that, you know i can tell when our cro. Is. Up early. Or. If he happens to be over in london because, you know, something. Doesn't look right to him it's, right to my phone. Nice bright and early so they've definitely taken it, uh adopted, that. It's also allowed us to kind of create a common, vocabulary. One of the things about the silent reporting, was. You know we still had systems, calling things. A sales phase, versus a panel for the real estate group and so we kind of all, come together, and, and there are things that. Now are just common, we. Brought those. That vocabulary. Together. I think. Always had. Measurements. Um. Kind of out, in the, company. But, uh this allowed, everyone to, you know really focus on what they want to measure. And. The the kpis. Were, you know everyone, knew what they were and they knew where they could find the measurements. And, i think that was. That was powerful, because sometimes, i think, we would just lose sight of what was expected. Um. And, i think that was really helpful. Um. There's something that we that our sales team likes to call visual performance, management. Um and of course, the data. Drives, that. But the. Visualization. Tool. Is the, obviously, the visual part of that. But you know drives. It drives some competition. Um helps, look for, trends and outliers. Um that, we they may not have seen in a spreadsheet. Before. So i think that's been, really powerful. You know we've, been able to introduce, third-party. Sources, into the warehouse, to help you know do things in the past we're really manual, and. Maybe, you know quarterly. Or. You know every once in a while, people get it. One of the things here that we're, working toward on the media spend side is. Is understanding. You know how we can. Identify. People who are spending. Media dollars, but not necessarily in that home industry. And can we go look at peers. Uh, of. Certain companies that are more favorable to us. And create, a story. As a media partner and say hey you know we see you spending, in these areas. You know. Can we show you that you know newspapers, and magazines. The circulation. Is way down, maybe you should move some of your dollars, you're not going to change your marketing, budget but your dollars can move into the at home and we feel like we're more, more. Effective. In that space so really just have those conversations. As a media partner. And we can do that with the, with the. Adw, now. Better to provide. Data to the business, than trying to wrangle that data and. Yeah, a common language, so let's just spend a couple minutes, on. Lessons. Learned, if you would. What, how can we help the audience kind of as, on their journey. Yeah so i mean i think one of the things. You know the big league of faith that we took was you know, let's go with oracle.
Um Let's utilize, existing. Uh expertise. So i think that was really. That was a good. Decision, on our part to say look you know. We know we can deliver. In this environment, with this technology. Staff that we have. So it allowed us to rather than focusing on the infrastructure. Focused on the data right so we can bring it, get it in there. Um. Get it modeled. And and then use visualization. And that really helped us become a business partner, so before we were kind of. The standoffish. I.t group that you know. Just waited for things to, bad things to happen, or, some fire drill requests. From. From a sea level, we need data, and you know we would all run around, for. Hours upon hours and spend weekends, wrangling. Things. So, you know by partnering, we we don't get surprised, by that as often as we used to. Not say we don't ever get surprised by a request, but. I think what it has done is. You know the things that they need to know they need to know. Um, and. We've put it out there in a way that they can consume, it relatively, easily. And what it's sparking, is other questions. Outside, of probably the normal. You know just how i need this or i need that. We're able to partner, and, i think. By having the quick wins up front. Not trying to try to deliver an entire platform, at once. We were able to, kind of. Organically. Ensure adoption. And you know the sales organization, adopted, very quickly. Um. And then the finance and real estate organizations. Came on and they were like hey, we want that you know we see what you're doing. Um and we were able to really come up with some creative, ways, to uh, not only, you know give them something. Some. Data that they weren't, getting before, but. We actually. Reduced, their reliance, on some kind of third. Party, or. Internal, applications. That they were using. And and we've shifted all that to the warehouse. Because the, cloud scalable. And economical. You know we can, we can try stuff i mean we can, my team might i tell them we don't have to be perfect, you know you have to. I want them to go try. Fail or succeed, and. Um i think, you know we've, especially in the last two years. Within, opening up of the tool sets we've done a lot, a lot, of, trying, failing. Something, doesn't work we can. You know we can abandon, it in the cloud and we can start over. And they don't feel like i've burned.
Cpus. Or, i, like bought a machine i didn't need. That. Well so uh, you know look thank you for, uh. Uh. Working with us uh sounds like the partnership. Is is really having an impact where you can focus on data and a data provider and partnering, with the business. Um, we'll help you helping you along with that journey. Um again thanks for being with us for a few minutes on on your story, and lessons, learned. And we're looking forward to greater things with, out front media in the future, thank you, scott look forward to talking to you again. Thanks. All right thanks derek, sorry to cut you off there but let's let's step it up a little bit um i'd like to introduce kim jackson, who is the vice president, of. Business, analytics, for adventist, health, kim welcome to the program. Thank you for having me. Absolutely. So uh let's get right to it can you tell us a little bit about adventist health and what's going on there, sure, absolutely, adventist, health is a faith-based, non-profit. Integrated, health system, we're in california. Oregon and hawaii, with over. 80, communities. We practice in the seventh day adventist, space. We service rural and urban communities, primarily, rural, our mission, is living god's love by inspiring, health wholeness and hope, 21, hospitals, medical groups and health plans with 28. 000 employees, and 1200. Physicians. Thanks for that i imagine this is really important. Given the, current environment, um. Now this is about data, so, and being data driven so walk us through real quick your values, on being data driven, absolutely, as a as a faith-based, organization, our mission and values, are extremely. Important to us and we, live by these through b statements. Two in particular, that reach out to me, uh within the data world is to be a mission owner you'll see through the rest of my organization. That we outsource. Most of our data strategy, and the work that we do but to be a mission owner we need to understand, your strategy, you need to drive your organization. So i take that to heart and we wanted to, have a new direction, for our organization. And, also. We needed to treat data as an intellectual. Property, and to do so. We needed to again, own our data. Use it appropriately, and help drive our organization. Forward. Perfect yeah outside of your people, your intellectual, property, data as andy put down the beginning is probably your most important asset so. We're about. You know, a little over halfway through 2020.. So what are your can you take us like a minute through going through your priorities, and i'd like you to touch on. Uh data literacy, if you would um you wrote an article. Um, recently, in, uh cio, periodical. About data literacy, and how that, how that's one of your priorities you walk us through that just for a minute. Absolutely, so i love to share what we're doing i know when i uh attend these events i want to see what my peers are doing in particular. In these areas. So i talked about how we outsource, so one of our first goal was to in-source, our analytic, team. And our leadership, i was one of those decisions, so i've just come on my anniversary, i've been with the organization, for about a year. To do that also, we had to, groom, our talent, and increase. The work that we do so that we were less reactive, and more proactive, so that we could help our organization. Data literacy, is a core fundamental, to the, fundamental, strategy, towards that, um so instead of us just being technical people, we really need to understand, our data domains, in the healthcare, space, understanding. The difference, between. Charges, and billings, and our payers, and what clinical documentation. In pharmacy, and laboratory. Health care data is super super wide. With all the traditional, business functions, as well as a really large. Um clinical, longitudinal, record, from the outpatient, the clinic it can go on and on and on so you can imagine, how, our, clinicians, at the bedside, our financial.
Uh. Fos. Don't have time to learn all of these data structures, that they need us to be experts, so that we can give them the insights, instead of them learning, a super complicated, data model, in data literacy, and having. Our, our, talent and our employees, understand, that so that they can deliver that service is essential. For us to be successful. Sensing a theme of partnering with the business not just being i.t people, so. Uh, rewind, a little bit um you said you've been with the with advanced health probably, at health for about a year, and what was the situation, going to the next slide. And what was the situation, similar to what kind of derek was talking about out front what was going on, um. When you uh when you dropped it, absolutely, so not only, um had we outsourced, our analytical, leadership, but we had outsourced, our platforms. Um our platform, was end of life and we were, with our vendor our emr, electronic, medical, record vendor. It happens to be cerner they supported. And we had our data warehouse. In their environment. So it was into life it hadn't had patches. In two years. We. We weren't able to own, our data model. We had a lot of problems, with stability. So we really needed to have a modern, environment. Our new strategy. Our 2030, strategy, isn't, dependent. On us being able to have. Technology, to support, a data model, support as hospitals, go from, just being hospital-based. To community-based. We needed more data sets and not just, hospital-centric. We needed to understand, the community, as well so we needed a broader, data sets and tools, and to help us do that, we had multiple, data platforms. Multiple, edw's. And we aren't able to connect, all of that data together, to actually, make it um, a, differential, in the work that we do, rising costs the cost that we were paying for this end of life. Edw. Was, very very large and we needed to reduce that in fact we were able to reduce it by 60. Um by moving so i'm real proud of that. And we were also. Dependent, on, our vendors. Roadmap. They had competing. Databases, they wanted us to, use their tools, and they made. Decisions, that wouldn't allow us to do what we were interested in doing in their same environment, so.
We Were really locking, key, um to their roadmap. Their desires, not our desires, as an organiza